ARTIFICIAL INTELLIGENCE/MACHINE LEARNING AS A TOOL FOR PROJECT

MANAGEMENT ENHANCEMENT: A QUALITATIVE STUDY ON THE FUTURE OF

TECHNOLOGY IN INCREASING THE EFFECTIVENESS OF PROJECT MANAGERS

by

LC Brooks Williams

AHMAD MOSTAFA, PhD, Faculty Mentor and Chair

KONDO LITCHMORE, PhD, Committee Member

MYLES VOGEL, EdD, Committee Member

Cheryl Boncuore, PhD, Interim Dean

School of Business, Technology, and Health Care Administration

A Dissertation Presented in Partial Fulfillment

Of the Requirements for the Degree

Doctor of Information Technology

Capella University

February 2024
© LC Brooks Williams, 2024
 Abstract

Projects are critical for all organizations. Over the years, a large number of literatures have

highlighted that most projects fail due to inefficiencies in the current process. Technology has

always been seen as a solution that could help in tackling the inefficiencies in the system.

Machine learning-based systems have been used increasingly in different activities across

multiple industries with the aim of increasing the efficiency of these activities while reducing

errors. Previous literature has also shown the capability and use of ML in project management

activities. At the same time, the literature on the topic is sparse, and there are no findings that

highlight the use and effectiveness of implementing these tools for project management. Though

the studies have highlighted the benefits and capabilities of its use in project management and its

potential, the lack of data in guidelines regarding its adoption is a critical issue. The studies in

the field have largely been aimed at exploring the possible benefits or barriers to adopting the

new technology within the realm of project management but lack the required guidelines that

would help adopt the tool and the critical issues that impact this adoption. The research question

for the study explores the perceived usefulness of AI/ML across industries within project

management. To find the answer to this, the study approach consisted of a generic qualitative

research design that makes use of structured interviews conducted with ten experts with prior

experience in using the tool for project management to express their perceived view of the tool

and its need. From the responses and thematic analysis, we found that ML benefits the

organization in risk management, resources management, communication, and planning and

helps improve the overall decision-making in different activities. That said, many factors have

been identified as responsible for the limited adoption. Some issues are the poor quality of data,
concerns about regulations on data use and security threats, lack of awareness, and concerns

regarding the cost, integration of technology with existing systems, and return on investment.

The responses also highlight the perception that collaboration and PMI developing guidelines

with the help of experts in the field would be critical for reducing the barriers. The research helps

highlight and provide more information regarding ML's role in PM activities and also helps

understand the limitations. It allows future researchers to further the development in the field,

providing specific areas that could be focused on.

 Dedication

First and foremost, I would like to give honor and thanks to my heavenly Father God for

providing health and strength to complete the tedious journey to possess a Doctorate level

degree. The effort used to complete this study is dedicated to my daughter and son, Lana and

Landyn Williams. Please always exceed the standard and challenge yourself to be the best at

what is important to you. I love you for eternity! The study is dedicated to the pursuit of

knowledge for future researchers. The journey was aided by the unwavering support of

colleagues, family and friends who have stood by me at every step of the way. Many thanks! The

final manuscript is the result of a lot of hard work, curiosity, research, and collaboration. May it

contribute and add to the growing human understanding and help explore and further the

research in the field.

iii
 Acknowledgments

Completing this Dissertation has been a challenging yet immensely rewarding journey,

and I extend my heartfelt gratitude to the esteemed members of the committee: Dr. Litchmore,

Dr. Mostafa, and Dr. Vogel. Their invaluable contributions have played a pivotal role in bringing

this project to fruition. I am honored to acknowledge their unwavering support, encouragement,

and insightful perspectives. A special debt of gratitude is owed to my Capella University

academic advisors for their continuous guidance, encouragement, and constructive feedback.

Their expertise and mentorship have been instrumental in shaping the research direction and

elevating the overall quality of the thesis.

I would like to express my sincere thanks to my Capella University colleagues for their

valuable input, critical perspectives, and thoughtful suggestions. Their contributions have

significantly contributed to refining the content and strengthening the overall quality of the

Dissertation. I deeply appreciate the interview participants who willingly shared their insights

and experiences, enriching the study. Their openness to engage in candid conversations has

added depth to the research. I want to seize this opportunity to express gratitude to the Capella

University academic community for creating a conducive learning environment that fosters

exploration and collaboration. The collective exchange of ideas and intellectual discourse within

this community has undeniably influenced the concepts presented in the Dissertation.

iii
 Table of Contents

Acknowledgments...................................................................................................vi

List of Tables ...........................................................................................................x

List of Figures .........................................................................................................xi

CHAPTER 1. INTRODUCTION ........................................................................................1

Introduction..............................................................................................................1

Background ..............................................................................................................2

Business Technical Problem ....................................................................................7

Research Purpose ...................................................................................................11

Research Questions ................................................................................................13

Rationale ................................................................................................................14

Conceptual Framework ..........................................................................................16

Significance............................................................................................................19

Definition of Terms................................................................................................21

Assumptions and Limitations ................................................................................21

Organization for Remainder of Study ....................................................................22

CHAPTER 2. LITERATURE REVIEW ...........................................................................24

Introduction............................................................................................................24

Methods of Searching ............................................................................................26

Conceptual Foundations.........................................................................................27

Project Management- Project Success and Life Cycle ..........................................29

Agile Project Management.....................................................................................36

Data-driven decision making .................................................................................41

iii
 Machine Learning ..................................................................................................44

AI, ML and Project Management ..........................................................................51

Synthesis of Research Findings .............................................................................67

Conclusion .............................................................................................................69

CHAPTER 3. METHODOLOGY .....................................................................................70

Introduction............................................................................................................70

Design and Methodology .......................................................................................70

Participants.............................................................................................................73

Setting... .................................................................................................................74

Analysis of Research Questions.............................................................................75

Test Run .................................................................................................................78

Credibility and Dependability ................................................................................79

Data Collection ......................................................................................................80

Data Analysis .........................................................................................................80

Ethical Considerations ...........................................................................................84

Summary ................................................................................................................85

CHAPTER 4. RESULTS ...................................................................................................87

Introduction............................................................................................................87

Data Collection Results..........................................................................................87

Data Analysis and Results .....................................................................................90

Summary ..............................................................................................................119

CHAPTER 5. DISCUSSION, IMPLICATIONS, RECOMMENDATIONS ..................120

Introduction..........................................................................................................120

iii
 Evaluation of Research Questions .......................................................................120

Fulfillment of Research Purpose..........................................................................125

Contribution to Business Technical Problem ......................................................134

Recommendations for Further Research..............................................................135

Conclusions ..........................................................................................................136

REFERENCES ................................................................................................................137

APPENDIX A. INTERVIEW GUIDE ............................................................................152

iii
 List of Tables

Table 1. Questions and Frameworks link ..........................................................................76

Table 2. Participant Demographics ....................................................................................87

Table 3. Codes and Themes ..............................................................................................91

Table 4. Theme 2- Responses ............................................................................................97

Table 5. Limitations mentioned across interviews..........................................................105

Table 6. Benefits of ML in Project Management............................................................111

iii
 List of Figures

Figure 1. Success of projects around the world ...................................................................8

Figure 2. Extended TAM model (Chismar & Wiley-Patton, 2003) ..................................18

Figure 3. Project Management Cycles ...............................................................................32

Figure 4. Iron Triangle (Dhillon, 2018) .............................................................................36

Figure 5. Waterfall model (AltexSoft, 2016).....................................................................37

Figure 6. Project success based on methodology (Ambler, 2018).....................................40

Figure 7. ML Techniques (Sarker, 2021)...........................................................................45

Figure 8. Figure 8 AI-Controlled Agile Project (Dam et al., 2019)...................................59

Figure 9. Barriers to adoption (IPMA, 2020) ....................................................................62

Figure 10. Seven Steps in Thematic Analysis (KADIR et al., 2021) ................................82

Figure 11. Commonalities and interlinks between the themes ..........................................92

iii
 CHAPTER 1. INTRODUCTION

Introduction

The increasing significance and intricacy of projects and their management have become

evident over the years, which has underscored the necessity for enhanced decision-making

capabilities to ensure operational efficiency and the successful attainment of project objectives.

The growth in digital data also means that project managers could use data from different

sources to help provide more informed decisions that could help improve the overall quality of

decision-making. Machine learning provides the capability to assess and analyze a significant

amount of data that could be used to identify trends within the data and provide analytical

answers that would help provide solutions to existing issues. This data could then be converted to

meaningful information through multiple machine-learning tools and adopted into decision-

making. Though machine learning cannot control project management, it is said to be a capable

support system that would help in adding value and providing improved decision-making

capabilities.

It has been mentioned that projects have failed at large due to poor execution, poor

management, or communication (Alami, 2016; Gulla, 2011). The growth in technology acts as a

buffer that could help tackle this issue. The research can provide the necessary support with the

help of technology across different aspects or elements of the project and is not restricted to just

one part of the project. The use of technology in tackling these issues is considered to be highly

essential for the future. Even with the multiple benefits identified for the research, the adoption

rate for the new technologies, especially in a field like project management that could benefit

from the analytical capabilities of the tool, has been limited . Understanding how the processes

1
differ and the limitations that influence their adoption would help develop platforms or

frameworks to improve the system's adaptation capability. The chapter explores the situation, the

research gap, and the need for the study while also establishing the critical research question.

Background

Computing technology and its evolution have allowed for improvements, enhanced the

quality of life, and provided more convenience for all processes that would often require a lot of

human effort and time. Artificial intelligence is one such technology that has significantly

improved in reducing effort, time, and automation, especially when it comes to processing or

analyzing large volumes of data and using the findings to make critical decisions. Artificial

intelligence has been used in automating and decision-making in different industries (Steels & de

Mantaras, 2018). Project management is also one such area where that has been growing

importance and presence of technology. The technological evolution has seen the increased use

of software-based solutions helping with various tasks considered high in time consumption and

cost. These processes or functions in project management include communication, information

gathering, data analysis, monitoring the overall strategy, and others (Prasad & Vijaya Saradhi,

2019). The focus of AI-based technologies has been to transform mundane and repetitive tasks.

While many advancements can be seen with project management and the use of AI, the

technology and its use are still evolving (Munir, 2019). To understand the role of AI and how it

influences the decision-making process understanding big data and digitization that controls the

processes and activities in the different industries is critical.

With digitization, a new standard has been set in the modern era where the amount of

data that is used for analysis and the eventual decision-making process has increased

significantly over the years. The term "big data" is often used to refer to large datasets that are

2
notable for their substantial size, diverse origins, fast generation rate, multiple formats, and

inherent worth (Hashem et al., 2015). This information is critical for businesses in their decision-

making process and holds different patterns that could help define expectations or measure

performance. This hunt for information is called data mining, which is extracting previously

unknown and valuable data (Witten et al., 2016). The raw data collected is mainly unusable and

would present no helpful information. Clever tools are often required to extract this information

from the raw data, which would help identify the patterns. AI is one such tool that can interpret

the data, learn from it, and use the said finding to solve the problem by simulating the human

brain functions. For example, AI-based neural networks (machine learning) can read and analyze

data, leading to decisions that would resemble the decisions made by human intelligence (Witten

et al., 2016). AI is limited in terms of understanding the logical reasoning attributes, and the self -

correction and learning capabilities are very abstract. That said, AI is capable of mimicking

human intelligence to a certain extent that can be deemed useful. AI possesses the required

capabilities that allow it to carry out complex tasks within a complex environment without

requiring the user's guidance and has evolved to provide improved performance based on self-

learning from experience. Most of the machines used in AI are considered to be narrow in their

application and are focused on solving a specific problem. However, more comprehensive

coverage of different areas of interest could be covered using the technology referred to as

General AI. General AI is not in existence at present, and it is believed that it will take time to

evolve and would be seen to be used widely by 2040 or 2050 (Witten et al., 2016).

Machine learning is said to be a sub-branch of AI and provides the required technical

tools that allow data mining and are used in extracting the information from the raw data stored

in the database (Tiwari et al., 2018). The information that is stored could be used for different

3
purposes, and finding such historical patterns within data helps make better decisions or

predictions for the future. Practitioners want to utilize Machine Learning (ML) to analyze mining

results and extract meaningful outputs from data (Kuo et al., 2016; Marsland, 2014; Miklosik et

al., 2019; Pham & Afify, 2005; Robert, 2014). ML creates a learning system that allows

computers to make use of data to improve themselves and fine-tune their prediction or solution.

At the crossroads of statistics, computer science, and data science lies the realm of machine

learning. ML is said to use elements from each of these fields in processing the data so that it

would be able to detect and learn from the patterns and predict any future activities. There are

areas in which machine learning has been able to achieve a significantly high level of

performance compared to human performance. There are other tasks that are still effective,

though, when undertaken by humans. One example that has been highlighted is the advances in

image recognition, where machine learning has helped increase the accuracy to 96%, which is

higher than the accuracy achieved by a human (Hughes et al., 2017). While ML does not reach

the human-level intelligence described with AI, the ability of the technology or algorithm to

learn from the data increases the total number of functions and complexity. ML is capable of

carrying out different tasks of varying complexity which would be difficult using the standard

programs. The learning elements also ensure an adaptive system that is focused on continuous

improvement in the accuracy of results.

ML provides many benefits, and some of the top benefits from projects that could be

achieved would include (Attaran & Deb, 2018):

• Improved planning – with the help of machine learning skills, the project

managers would be able to develop plans, consider best practices, and schedule

the services in better flow while considering the past projects and where each was

4
 positive and where the projects struggled. The increased level of information

available allows for making improved decisions and thus improves the overall

planning and which steps need to be undertaken. The simulation capabilities

through data also allow the project manager to check multiple scenarios manually

or by ML algorithm suggestions that provide an ideal path.

• Improved services: Machine learning allows for tools that allow the capability to

provide more automated services like schedule planning or operation based on

existing information that is controlled through ML algorithms. This could involve

detailed reporting, communication with stakeholders, planning and progress

update, and many other services that were to be done manually in the past.

• Enhanced risk management: the capability to analyze data from different sources

and past risk situations allows one to understand the risks that could often be

unseen under a traditional manual approach. Machine learning allows for the

integration of external parameters like policy changes to be analyzed and the risk

sensitivity analysis to be undertaken based on these changes that provide better

decision-making capabilities. Also, machine learning allows for simulations of

different scenarios to identify the best ways to mitigate, avoid or eliminate the risk

most effectively.

• Cost and time savings: With the capability to integrate data from different sources

and analyze them, the number of people required for the process would be

reduced; at the same time, the automation of some operations or updates in

information makes the job faster and more accessible for the project manager.

5
 • Enhanced analytics: the amount of data that could be analyzed increases. Under

manual process, there are limitations in data that could be explored; in contrast, in

machine learning, even those data considered irrelevant could be integrated, and

patterns identified from those data add and provides a new dimension of

information.

• Increased revenue: With machine learning and better planning and decision-

making, the projects can be completed, and fast-tracking could be achieved,

allowing for adopting more projects that would help increase the overall revenue.

Exploring in-depth, the organization, based on a survey, believes that the use of ML

provides them with more extensive data insights which were previously unable, which allows

them to make more accurate decisions and also predict the future better (Mahdi et al., 2021;

Malik et al., 2021; Q. Wang, 2019; Wei & Rana, 2019). It provides a competitive advantage in

analyzing the behavior of its competitors and acts according to the market requirements (Attaran

& Deb, 2018). Machine learning provides faster data analysis, allowing faster transformation or

adaptation to new requirements. It improves R&D and reduces the costs involved in projects or

other activities. All these benefits have made technology highly attractive in various fields,

including project management, where AI and ML have been used increasingly in various

applications. Machine learning is said to be significantly limited in some aspects of the process.

The adoption of ML, though considered easy and capable of providing increased benefits, does

have significant limitations. These limitations could include factors like the process's complexity

and investment requirements that are not affordable to all (Barocas et al., 2020; Kopper, 2019;

Stewart, 2019). There are also no specific guidelines or measures defined or policies or standards

developed to implement machine learning into project management. People are also unaware of

6
the capabilities, and there is a lack of spread of information on these capabilities, and very little

research has been done in the area of project management that highlights the critical capabilities

of machine learning and project management. Most of the present papers are also theoretical and

mentioned based on theoretical assumptions and not on practice. It is vital to comprehend the

obstacles that hinder users from adopting machine learning despite its various benefits and to

analyze the causes of distrust in the system.

Business Technical Problem

The success of a project depends on completing multiple tasks within a set timetable,

maintaining a predetermined degree of quality, and keeping to a specific budget. The activities

within the projects are unique, challenging, and connected, requiring intricate management to

ensure the achievement of objectives. The success of a project is influenced by project

management, as projects are complex and interwoven series of distinct operations. The

successful execution of these tasks depends on the specific limitations of money and time and is

in accordance with the established scope of the project. Projects are often characterized by other

generalizable attributes like life cycle, uniqueness, conflict, and interdependencies. Project

management is the knowledge possessed by the project manager in planning, monitoring,

organizing, and controlling the different activities within a project and ensuring stakeholders are

involved and committed to the goals and objectives (Dahie et al., 2017). Project success and its

definition are challenging, though the commonly accepted norm for project success measure has

been defined based on the project achieving its objectives related to cost, time, and scope, which

brings in the limitation of the project being considered a success or failure only after it has been

completed.

7
Figure 1
Success of projects around the world

Note. Success rate of projects globally which would mean the rest have failed. Inspired by

original work “Success Rate by Location” by Consulancy.uk, 2020. Retrieved From

https://www.consultancy.uk/news/24677/most-construction-and-engineering-projects-are-

unsuccessful.

In a 2003 report, it was mentioned that for those companies that would not be in the top

25% of technology users, three out of 10 projects would fail on average (King, 2003). This

would mean nearly 30% of all IT projects fail. This would mean that 70% of the projects are a

success, but the measure of success and to what level these were successful is not something that

has been measured. The poor quality or lack of reliability of project management in

organizations has been attributed to being one of the main reasons for the failure of projects in

the 2009 report. The report also states that 25% of the projects are faced with outright failure,

while 50% require rework, and around 20% to 25% are said to provide no ROI (Arcidiacono,

8
2017). There are many reasons for failure within project management that have been highlighted

in studies, like lack of concrete strategies in planning and scope creation, poor or clear

stakeholder communication requirements, weak management, differing objectives or

expectations from the project, poor technology adoption, and lack of support from management.

That said, IT projects are not the only ones that fail. It has been mentioned that two-thirds

of all significant construction and engineering projects are known to fail. In construction, it was

found that 66% of the projects were often delivered past their time frame, and also, in most

cases, they were not able to meet the cost and quality expectations either. The success of

immense construction and engineering projects was just 42% in North America, while in Europe,

it was just 33% (Consultancy UK, 2020). This shows that project failure rates are pretty high,

irrespective of the industry. The findings in the literature highlight that different influencing

factors play a critical role in influencing the success of projects, especially those that are found to

be capital-intensive. It has been mentioned that when teams are made of highly skilled people,

the projects are 1.8 times more likely to succeed. It was also mentioned that projects with high

compliance with the process are two times more successful. Project success is also dependent on

when the project controls were used, and it was found that those projects that instigated the

control early in the project life cycle were 52% more successful. One crucial area that has been

said to be a critical player for project success, specifically in construction, is technology

adoption, like Business Information Modelling (Consultancy UK, 2020). BIM is a concept that

uses AI and ML technologies through a modeling process that provides professionals with better

insight and tools to plan, design, manage, and construct buildings more effectively and

efficiently. It has been mentioned that the organization that used BIM were 30% more likely to

9
achieve success, and those who used 4D planning had a success rate of 50% (Consultancy UK,

2020). AI and ML are in their infancy in project management and are slowly being adopted in

different projects. With the analysis, it has been highlighted that projects that undertake

automation would achieve significant improvement. As per the report, there would be a 79%

chance of meeting quality and a 61% chance of meeting the time objectives. It is also stated that

the projects were 2.4 times more likely to perform better within integrated working and project

control. Project Management consultancy firm CEO Bryn Lockett mentioned that AI and ML are

the future of complex and critical projects, and these technologies would be critical in ensuring

that the integration and success of the project is achieved (Consultancy UK, 2020).

Regarding critical decision-making, project managers are still considered critical as their

experience would help validate the choice. This is based on their intuition and experience; the

decisions could change. Though AI is capable of making quicker decisions, there are still

limitations to the technology as it is still in the initial stages of its development (Magaña

Martínez & Fernandez-Rodriguez, 2015). Thus, ML is most suitable as it would help in detecting

the required patterns in the data and allow for the development of suggestions or various

information based on the analysis requested that would aid in the decision-making process and

thus would leave the control in the hands of humans but provide them with more information

than they previously made use of.

The specific business technical problem is the reluctance seen when it comes to adopting

ML technology in project management even with many potential benefits highlighted for the

same (Kelepouris, 2023; Uysal, 2021). The study makes use of existing literature along with

interviews in highlighting what factors are limiting the adoption and what the perception is on

tackling them that would boost the use of ML and AI in project management and boost the

10
overall efficiency of project management across industries. There has been a significant push for

research and processes regarding adopting new technology to be added to processes. All the

failures in construction, engineering, and others similarly highlight the growing importance of

using advanced technology. With growing digitization and increased level of information

available with the technology, the use of Machine learning in project management is not

something that needs to be forced but adapted to. In the simplest form, it provides analyses and

information based on the analysis that allows for gaining additional insights and helps with

decision-making. There are also areas where the technology could help in automating the process

and use the technology to reduce the workloads, giving the supervisors the required time to focus

on other aspects. The capability that it provides in identifying risks that in the past were not

analyzed due to the complexity and time required in data analysis is eliminated through the

adoption of ML technology. That said, businesses are still being researched in the adoption. In

light of the increasing importance placed on sustainability and cost reduction, it is imperative to

identify and eliminate behaviors or causes that hinder the implementation of procedures.

Adopting this proactive strategy is crucial for improving the overall efficacy and efficiency of

project management. While the technology and its benefits have been presented and promoted,

there is still resistance to adoption. Understanding them and tackling this issue at the core would

help researchers tackle the issues and create the required assistance or frameworks to increase

adoption.

Research Purpose

The research examines the barriers that limit the adoption of ML and AI in project

management, even though it has been shown to improve project management's overall

effectiveness and efficiency. We understand the capabilities, and over the years, there has been

11
an increase in the number of studies that have been explored that are focused on using AI -based

technologies in project management and how they help project managers. That said, the role

played by machine learning has been limited to studies focused on specific activities or

applications that are carried out using ML in different activities of the project. There is currently

a dearth of comprehensive investigations into the integration of machine learning (ML) and its

practical implementations in the field of project management. Furthermore, there is a lack of

comprehension among industry specialists regarding the potential and limitations of this

technology. Closing these knowledge gaps is crucial for advancing research in project

management and promoting the integration of technology to fulfill project needs. The research

would help identify areas of weaknesses where further research is required or innovations need

to be undertaken and help with its progress for the future.

To get the required answers to the research questions, the research would undertake a

two-method approach with triangulation of the results in providing the answers. There are certain

aspects of the research, including the development of the interview tool and also to compare

better the findings regarding the challenges or ML adoption that has been mentioned for other

industries or other processes, which can be then compared with the help of interview to highlight

how they are different in project management or how they are similar. Considering that the focus

on the limitations is an area that has been least explored, especially when we specify or focus on

the construction industry. While it would help provide context regarding the situation and the

challenges for its adoption, specific perspectives based on real-time experiences cannot be

achieved from literature, which would mean we would need to rely on assumptions. This is also

because even if the literature is an interview-based study, there could be an inherent bias in how

the results have been presented, which could significantly influence the situation. The interview

12
helps counter this challenge and allows us to understand better the involvement of machine

learning technology in project management and what limitations the people who have worked in

project management have seen that would highlight a specific problem related to the area being

explored, thus gaining critical information. This information or details on the limitations could

then be tied up with literature to give a complete idea of the situation and help identify areas

where more needs to be done.

Through this methodology, the research would help highlight the factors or limitations

that would need to be undertaken in future research to find an ideal solution or provide initial

recommendations for sorting these issues in organizations. These problems and creating a critical

framework or solution to adoption and having the companies be more aware of the need for ML

technology in project management is critical.

Research Questions

While exploring the research topic, it was clear that elements have not yet been explored

in the literature. ML has grown in popularity, and research on the topic and field has also seen

increased relevance and importance in recent years. However, the research available on the topic

and understanding why businesses, especially in an industry like the construction sector where

the failure rate of projects is so high, fail to adopt the technology and adapt them in the process

would provide them an advantage. Understanding the concerns is important as it would allow for

a better understanding of how the adoption of technology would be possible, and this is critical

to improving the project management process and translating the theoretical benefits and, in

some cases, the real benefits that could be gained from the technology adoption that are explored

in literature. Identifying obstacles to adoption and identifying areas that need further

13
investigation are essential measures to ensure general acceptance and uniformity. Given this

information, the primary research question that arises is as follows:

What is the perceived usefulness of AI/ML across industries in project management?

This question would help explore the challenges faced with adopting machine learning

and why the technology has not been able to attract adoption even though it promises many

benefits. The question would help understand its evolutions and uses and highlight which areas,

as per industry experts and people that have used these systems, need improvement or act as a

hindrance in their adoption.

Rationale

The research question aims to understand perceived usefulness of adopting ML

technology. Understanding why adopting the technology is critical for the future is critical. The

existing literature points out the need for improving project management techniques in different

industries to improve the project success rate. Companies are focused on efficiencies, and over

the years, many industries have adopted the latest technologies. However, not all of these

organizations are happy when it comes to the adoption of new technologies. There are many

reasons for this fear, and understanding this is critical. While the technological adoption of ML

and the dears of the technology adoption have been researched, there is a lack of clarity or

understanding when it comes to the project management context, especially in the construction

industry. Each situation is unique, and each industry's requirement makes the concepts and

challenges unique, and understanding these differences would help identify the factors.

With the growing uncertainties seen regularly and the scarcity of natural resources, there

is an increasing need to address sustainability issues in every aspect of business life, including

projects. Sustainability is critical in the integration of finances. Nature and social responsibility

14
factors are responsible for harnessing the present resources judiciously and ensuring the

availability of the same resources for future generations (Chawla et al., 2018). This means there

is an increased focus on ensuring there is less wastage of resources in all activities, including

project management. The traditional technique is said to exploit the resources in seeking the

optimal combination of time, quality, and cost performance and to increase the overall benefits

for the stakeholders. The approach is considered reductionist, has a significant impact, and

creates sustainability challenges. The increasing push for sustainable practices in the projects has

increasingly led to the need for different approaches that would help improve the overall

management process (Armenia et al., 2019; Costantino et al., 2015). This is one of the areas

where advanced technologies could provide an edge and help achieve these goals.

It has been highlighted that in a project-oriented industry like construction, data has

become an integral part of its operation, allowing them to get better information that helps them

in their decision-making (Mahajan, 2021). The use of data has been claimed to help improve risk

prediction, contractor management, safety training, and predictive modeling. Big data has been

seen to help improve the overall quality performance and enhance the technical capability that

would act as a mediating factor between adaptability and quality performance (Sang et al., 2021;

Torrecilla-García et al., 2021). With extensive data being available for analysis, there is also the

need for techniques and technology to extract unique and vital information from these data to

help make the right decisions. The adoption of ML and big data is relatively new. These

technologies allow for the transformation or shift of project management to agile project

management, which improves the flexibility and adaptability to the requirements and changes in

the dynamic environment we presently have. Though many studies have explored the adoption of

the technology-based technique in general, there is a lack of studies that focuses on identifying

15
best practices and limitations of these methods and how they could be used in different projects,

along with the perspective of people who have used these tools and their experiences.

Conceptual Framework

The research explores or focuses on understanding the adoption of ML practices in

project management. This study investigates the process of incorporating new technology into a

practice that normally does not use technology. It analyzes the elements that affect the

acceptability of this technology in this specific context. When a new technology is introduced,

multiple factors influence its adoption. Even in healthcare, like vaccines, specific guidelines

influence the adoption, and not all technologies would be widely accepted. The lack of adoption

or acceptance could be studied to understand why the individuals are unwilling to test the system

or product and make the required changes or parameters to help increase adoption. Businesses

need to be aware of the factors limiting the adoption of their products, which would allow them

to develop strategies (Surendran, 2012). These strategies could be aimed at developing new

products or could be focused on creating conditions that are more suitable or ideal for the

company. Without being aware of the challenges a technology faces and why it is not accepted,

the products could fail, or technologies that could be beneficial could be ignored as people would

consider them unsuitable. The perception is governed or influenced by those who use the

technology and their perception. Understanding their views on the new technology and why they

feel a particular technology is not suited or cannot be accepted helps develop a change that

would tackle the identified issues to improve the technology acceptance.

This study, like many others before, focuses on the phenomenon related to the adoption

of new technologies and what influences their adoption. A model that is widely used to

understand the perception of the users regarding the technology that is adopted is the Technology

16
Acceptance Model (TAM) (Hu et al., 2019; Rahimi et al., 2018; Surendran, 2012; Vuković et al.,

2019). Over the years, this model has been used to understand the person’s intention behind a

particular behavior while using technology, predicting the use and acceptance of the technology

by companies and individual users. The TAM framework has become an increasingly popular

tool focused on understanding technology acceptance. The framework outlines the expected

results, effort considerations, social aspects, and enabling conditions necessary for a deeper

understanding of the adoption process or influences (Rahimi et al., 2018).

The basis of this inquiry involves analyzing the development of AI systems, including

their subsets, such as ML and deep learning, in the field of project management. The study

seeks to clarify the influence of these technologies on elements that contribute to success and

highlight the growing significance of data analytics in attaining sustainability goals in various

projects. In addition, the study examines the use of agile project methodology, which improves

flexibility and adaptability, therefore adding to the transformative elements of projects. The three

concepts of success, sustainability, and agile are focused on due to the different focuses of the

three concepts and their importance in project management in the current dynamic environment.

TAM focuses on two critical factors in the user’s decision-making about accepting this system

by the potential users. These two factors are perceived ease of use and perceived usefulness.

Davis and his colleagues 1992 developed the framework to explain how individuals make the

decisions to accept new technologies and use of them (Naeini & Krishnan, 2012). The

developers believed that the perceived usefulness would highlight how the people believed the

technology would help them enhance their performance.

17
Figure 2

Extended TAM Model (Chismar & Wiley-Patton, 2003)

Note. Provides additional elements of TAM that adds a comprehensive outlook on the

parameters. Inspired by the original work “Extended TAM Model (TAM2)” by Chismar and

Wiley-Patton (2003). Retrieved from https://ieeexplore.ieee.org/document/1174354.

Within the scope of this study, an expanded version of the TAM2 model encompassing

cognitive, instrumental, and social influence processes has been introduced . This framework

18
incorporates four cognitive factors, namely job relevance, output quality, results, and perceived

ease of use, all of which contribute to shaping the perception of perceived usefulness.

Simultaneously, three social influence processes—subjective norms, voluntariness, and image—

shape the overall perception. Job relevance, in particular, represents an individual's perception of

the technology's applicability to their specific job (Chismar & Wiley-Patton, 2003). The

assessment of the system's capacity to accomplish necessary tasks determines the level of output

quality, while the measurability of the outcomes associated with the technology's utilization is

referred to as result demonstrability. Subjective norm is individual perception of whom the

people feel is required to use or not use the technology and perceive use is the usefulness of the

technology in the process, which in this case would be the project management. Image refers to

how the use of the technology in this scenario would add to the organization’s competitive edge

among its competitors, and voluntariness refers to the extent to which the process adoption is

critical or could be considered mandatory or non-mandatory (Chismar & Wiley-Patton, 2003).

These factors are critical for the study and understanding the role that ML could play in helping

with the decision-making process in project management and hence is the ideal theoretical

framework choice.

Significance

This study aims to elucidate the effectiveness of integrating machine learning into project

management and its impact on project success. The investigation will extend to understanding

the existing constraints within these systems. By identifying the specific phases of the project life

cycle where machine learning proves most effective, the research endeavors to assist companies

in adopting these practices. Concurrently, the study seeks to pinpoint limitations or areas of

concern, thereby guiding further innovation and research in the field. It will allow us to analyze

19
if ML or deep learning techniques are more compatible and effective than AI. This would allow

companies to identify suitable methods or technologies and be prepared to face some of the

challenges they could face from the deployment. They would also know how the technology

could be used for maximum benefit.

While the project and its significance if large for the field of project management, the use

of TAM in ML-based adoptions could help in developing frameworks or understanding the

limitations of the existing systems and taking the required measures in developing models that

are required for adoption and also testing common concerns in other areas of the implementation.

The adoption concerns and challenges faced could be compared with other adoptions in different

fields. To enhance comprehension of the situation, it would be beneficial to emphasize common

concerns associated with the adoption of machine learning practices. ML and its adoption have

become increasingly critical and with the focus on efficiency, developing a standard protocol that

could then be modified based on the requirements is extremely critical. ML-based adoption could

also pave the way for other technology adoption in project management. This could include deep

learning or other techniques and integrating multiple tools to automate the process to a certain

extent. The study would also help highlight the limitations of the existing ML technologies and

what additional changes or developments are required to ensure better adoption and better

services. Understanding the concerns faced in ML could be used to test other tools; this would

allow researchers to understand better users’ behavior based on different technologies and their

use of information. Once this model is developed, researchers could also focus on developing

adoption models that are industry specific, as the project management is said to differ between

industries while having a common base from which the development could be initiated.

20
 Definition of Terms

Agile Project Management: Agile Project Management is an iterative approach to project

management that breaks the entire process into smaller groups that would need to be completed

separately (Arefazar et al., 2019; Fagarasan et al., 2021). This technique increases stakeholder

communication and collaboration, providing better control, continuous improvement, reduced

risk, and enhanced flexibility.

Artificial Intelligence: AI is the capability of computers to make decisions like humans

based on the data available at hand, with the capability to learn and improve.

BIM: Building information modeling is a process used in the construction where different

data from different sources are used in developing a 3D model analysis of construction projects

for effective management and risk assessment (Deng et al., 2019; Torrecilla-García et al., 2021).

Deep Learning: A type of AI considered to be a subset of ML that uses artificial neural

networks capable of mimicking the learning process of human brains and can be used for

unstructured data analysis (Elboq et al., 2021; Tiwari et al., 2018).

Machine Learning: Algorithm used in data analytics to help in identifying patterns and

improve decision making (Pham & Afify, 2005).

Project Management (PM): The use of specific tools, skills, knowledge, and techniques

focused on leading a group or team to achieve specific goals set for the projects within a given

constraint (Project Management Institute, 2017).

Project Success: The project’s success is based on scope, time, and cost dimensions

(Pham & Afify, 2005).

21
 Assumptions and Limitations

Two primary industries that face significant project failures are IT (Software

development) and construction, and hence exploring the effective use of the technology within

these industries is assumed to help tackle similar problems within other industries at a certain

level. It is also assumed that the case studies or literature review-based exploration would shed

details that would provide more insight into the topic, and interviews are not biased.

Understanding the different tools used in various stages of the project lifecycle and how they

could be effective at different levels would shed more light on how ML could be used. Since

Machine Learning is a branch of AI, we would consider the studies in AI and Deep Learning for

data collection.

Individual preferences could be a part of the study as there would be people who are

favorable to adopting technology, and there are also those that would be against adopting these

technologies within project management, which could significantly limit having a clear

understanding of the benefits or challenges. The responses might be biased and have individual

opinions that would contradict reality and require further quantitative or experimental analysis,

which is not a part of this study. The limitation in time and willingness of participants across the

two industries to be a part of the research due to unwillingness or the time required is a

significant limitation that would hinder the total number of participants.

Organization for Remainder of Study

The paper will be structured across five chapters. In the initial chapter, emphasis will be

placed on comprehending the background and the problem under investigation. This section will

establish the research questions, delve into the significance and rationale behind the study, and

lay the groundwork for the theoretical framework. The second chapter is the literature review

22
focused on one exploring critical concepts regarding machine learning, project management, the

use of big data and analytics in project management, agile methodology and its relevance, and

the growing importance of AI and subset technologies in project management. The third chapter

is Methodology, where the research philosophy, methods, and tools used in data collection will

be explored, along with the techniques used to analyze and validate the study. Chapter four is the

results where the data collected would be presented after analysis. The final chapter is where the

findings of the study in the previous chapters are interpreted, recommendations are provided for

organizations and future research, and the paper is summarized or concluded.

23
 CHAPTER 2. LITERATURE REVIEW

Introduction

Artificial Intelligence-based Project Management is considered an integrated system

capable of undertaking project administration without requiring input. AI allows insights into

different activities, provides recommendations based on collected data, and helps in decision-

making (Tambe et al., 2019). It is said that the evolution of AI is promising, and soon, they will

have the capability to match the right kind of resources with the right kind of role. For example,

in the hiring process, HR managers are investing more in AI technology to improve the

efficiency of the process. It has been found that recruitment has seen a 20% improvement

through this process (Chou et al., 2015). The findings also reported improvements observed in

revenue, which was around 4 percent, while the turnover rate, which used to be a significant

challenge for the organization, was reduced by 35%.

It is important to mention that AI also has the capacity to reduce the workload of project

managers. It can serve as a mechanism to understand the work rhythm of different team

members, enabling better distribution of work depending on individual abilities. Consequently,

this has the capacity to improve the total standard of work. The technology is believed to be

capable of aggregating workplace behavioral patterns and creating a centralized knowledge

system for workers that would help improve the overall quality of the project and its consistency

(Conforto et al., 2016). Project management AI could be used to aggregate knowledge and

behavioral patterns. It could study each worker's different skill levels and capabilities and allow

for allocating these workers to the right job. It also ensures that the information is all stored and

reinvention is prevented. This concept is still theoretical but shows AI’s capabilities in the field

(Tambe et al., 2019).

24
 Data-based analytics and scanning of the environment allow working in different

conditions and allow AI-based technology to detect invisible warning signs that could

significantly risk or impact the projects. Taking the example of a construction project, AI would

be able to see the performance of the equipment with regards to its wear and tear, change in

employee facial expressions, measure the quality of air, or identify unsafe work conditions,

which could help create warning and alerts that would enhance the safety (Costantino et al.,

2015). It is also tough to identify variables or factors that could be used in measuring quality and

accuracy. This is a tedious process that requires many hours, is tiring, and could significantly

impact the project manager. At the same time, ML does not get tired and could use the more

significant amount of data, including past projects, to identify patterns and provide suggestions.

ML and associated technologies are still in their infancy, and their adoption in project

management is still being explored; hence, there are limited tools. To understand how it

enhances project management, it is critical to understand the various concepts like project life

cycle, project success, agile and waterfall methodology, and the role of data in projects.

The research centers on comprehending the potential of ML in the realm of project

management. Additionally, it aims to identify the factors constraining its widespread adoption

within the construction and IT industries. These industries, characterized by their project-

oriented nature, present distinct requirements and processes, making the study particularly

relevant to understanding the nuanced applications and limitations of ML in diverse project

environments. In order to understand the definition of project success and project management,

the study first delves into these topics, which would help in understanding the different stages

and life cycles and how project management works in these stages. Once we have this, we look

at the concept of agile project management, a technique used for software development that

25
provides flexibility and its adoption in different industries. We also look at how data is used in

decision-making in various fields, including project management, and then explore AI and ML-

based practices. AI is explored to show the variety of tools and techniques adopted in the project

management practice. Adopting AI practices could also help better understand the capabilities

and benefits gained while exploring ML-based adoption could be areas where it has been used in

project management. In addition, we look at how the use or adoption of AI has been introduced

as a response to the limitations identified for agile project management and the benefits gained

from the said adoption. The literature thus would provide an overview of the different aspects of

ML technology and AI in general and help understand AI’s preferences. At the same time, ML is

not as widely adopted, which would help develop the questionnaire.

Methods of Searching

To identify appropriate journals for the topic, the research leveraged three databases:

Google Scholar, Science Open, and CORE. These databases are renowned for housing an

extensive collection of studies published over the years, offering convenient accessibility to a

wealth of academic literature. The search was carried out independently across the three

databases, and any study found repetitive was removed from the files used in undertaking the

literature review. Based on the structure mentioned in the previous section, the first search term

used was project management cycle and project management success. Studies with these key

terms in the title and body of the articles were selected. It was focused on ensuring that most

studies were not older than 2010, while for the project success factors, due to the study’s relative

limitations, the time frame was reduced to the 2000s. Following this, other keywords explored

included Agile Project Management, big data in decision-making in project management, AI in

project management, and ML in project management. The study ensured that all articles were in

26
English and only those that satisfied or provided information on the topic and were relevant to

understand the role were selected. Any non-English paper was removed from the search. Any

paper that was opinion papers or magazine articles, or web pages were also not used.

The search was performed with a specific set of terms, which included "Project

management cycle," "Project management success," "Project management and AI or ML,"

"Barriers and challenges to technology in project management," and "Benefits of AI and ML in

project management." These keywords are representative of how the advanced search was used

in finding the articles based on the key terms and/or advanced functions.

Conceptual Foundations

The underpinning or precursor to TAM can be traced back to the theory of reasoned

action. This conceptual model was formulated with a primary emphasis on predicting human

behavior in a broader context. Comprising two key elements—attitude toward behavior and

subjective norm—the theory identifies these as determinants influencing individual behavior

(Fayad & Paper, 2015). Attitude in this would refer to the personal feeling, positive or negative,

under the evaluative effect regarding the performance of the targeted process. The subjective

norm is the individual perception that most people consider an individual essential and believe

that particular behavior or action should not be performed. The Theory of reasoned action is said

to have played a critical role in developing and adopting the Technology acceptance model

(Fayad & Paper, 2015).

One of the most accepted models in research when it comes to exploring technology

adoption has been the TAM model. The model was first introduced in 1986 by Davis, and it has

since served as a model or framework that has been used to explain or predict user behavior

when it comes to new technology (Elshafey et al., 2020; Tantiponganant & Laksitamas, 2014).

27
TAM serves as a potent tool for understanding how external factors impact belief systems,

mindsets, and the motivation to use technology. It unveils two crucial cognitive perspectives: the

perceived significance and comprehension of ease of use. Essentially, the framework is based on

the idea that a user's acceptance of a technological system is influenced by important aspects

such as their intention to use it, their attitude towards it, their perception of its usefulness, and the

system's simplicity. These elements significantly influence the user's behavioral intentions,

attitudes, perceptions of utility, and perceptions of simplicity, both directly and indirectly. As per

the model, external influence would impact the product's intent and actual use, and the mediating

effects are found to be on the perceived utility and ease of use. The most critical determinants

when it comes to TAM are the perceived ease of use and usefulness. The model has been used in

other research, specifically in construction-related research. Some technologies like Building

information modeling and Augmented Reality have been explored using this model. The TAM

has also been used to derive factors that would help maximize the approachability of technology

acceptance models related to construction before implementing new IT-based systems in the

construction field.

TAM is a modified version that makes use of the perception influence when it comes to

the ease of use or usefulness to understand the individual and the intentions to use or predict the

usage behavior. This is limited, and the model used tries to expand on the current model and

adopt TAM2, which, like the standard framework, has been used extensively to explore

technology adoption. This has led to adopting more specific factors and how they influence the

usefulness of perception. One example of this is Job relevance which is looking into the

importance of the technology in the job or field, which is essential in this study as we are

exploring the limitations or factors that act against the adoption of ML. If people believe that

28
technology is irrelevant to the field, adoption would be difficult. Similarly, result demonstrability

is their experience and first-hand knowledge of the results of using technology. A positive

experience or result would positively impact usefulness, while the opposite would also be true.

Thus, these additional factors allow us to focus on specific aspects of behavior that influence the

usefulness factor to evaluate the technology better. TAM as a conceptual model has been capable

of finding technology acceptance in different settings and helps evaluate and understand the

resistance often seen against the adoption of a specific technology (Al-Ghamdi, 2009; Al-

Momani et al., 2016; Amadu et al., 2018; Çivril & Özkul, 2021; Dickson et al., 2021; Karim et

al., 2022; A. Kumar et al., 2020; H. H. Lee & Chang, 2011; Lu et al., 2003; Zhou et al., 2021),

making it the ideal model.

Project Management- Project Success and Life Cycle

Considered to be one of the most prominent schools in the management field, the

definition of project management would be as follows:

“The employment of various techniques, skills, and tools focused on developing the

project at hand and making use of them in achieving the objectives of the said project.” (Project

Management Institute, 2017)

Projects could be considered to be activities that are temporary in nature and are

undertaken with the aim of developing a set of unique products, services, and results that a

customer requires; thus, each of the projects is considered to be unique. The project needs to

ensure a point where it would start and a point where it ends, along with having clearly defined

objectives or goals along with a defined budget in achieving the scope of work described while

also ensuring that the quality and performance requirements are also met. When project is

29
considered as a problem that needs to have a solution it could lead to further complications and

challenges (Heagney, 2012). Project Management is said to be accomplished through a group of

42 logical processes that are said to be grouped into five process groups: initiation, planning,

execution, monitoring and control, and closing. The project requirements would be inclusive of

the time cost scope time constraints.

In this process, the project manager would be the enable that would be focused on

completing the work, running required interferences for the teams, and buffering or supporting

them from the various external or even internal forces that could disrupt their work and hence the

project. It is imperative that the project manager act as a leader who also carries out certain small

administrative parts of the project like scheduling and control. Without proper leadership, the

projects will not be able to achieve their primary objectives (Heagney, 2012). There is often a

misconception among people who think project management to be just a project schedule which

is far from the truth. Though scheduling is critical, it is not the most crucial aspect of managing

projects. It is critical to developing a clear understanding of what the projects are supposed to

accomplish, breaking down the different activities into small and achievable groups, and

assigning the right team to these works to ensure high quality and reliability.

Over the years, multiple models have been developed. One such model is the model

introduced by Mike Bell, which considers five elements: risks, project, outputs, scope, and inputs

(Johnson, 2013). The project's scope establishes limits, inputs conform to requirements, and the

project is segmented into five stages, emphasizing results and deliverables; critically, proficient

project management relies on efficient personnel management. (Cooke-Davies, 2002).

Researchers have stated that the people who deliver the projects, their outcomes, and the systems

facilitate their work. It has been mentioned in the literature that failure or success would be

30
heavily influenced by the stakeholders whore a part of the project (Henrie & Sousa-Poza, 2005).

The observed transition suggests a break from traditional project management technology and

methods, refocusing attention on human and interpersonal factors, particularly in the context of

research and development projects (Leybourne, 2007). Within this framework, the focus is

mainly seen on the development of activities, making sure changes are facilitated , and, last but

not least, providing suggestions. When it comes to the critical element in the process, it is said to

be the willingness of the stakeholders involved, which are managers and employees, to embrace

the said changes and adapt to the recommendations made (Jetu & Riedl, 2012).

Project Management Cycle

The division of projects into multiple phases is undertaken to help simplify the process

and allows the leadership to move in a better direction. Based on the standard PMBOK literature,

there are five phases. The initiation phase is where the project objectives would be identified,

which could be based on a problem faced by the organization or an opportunity it has from the

changes. In this stage, a feasibility study would be undertaken to help investigate whether the

addressed options help achieve the objective and provide a final recommendation based on it

(Project Management Institute, 2017). Only after the project objectives are established is the

leader or manager appointed, and the support team for the project manager is also identified.

31
Figure 3

Project Management Cycles

Note: Various phases of project management cycle. Inspired by the original work “5 Phases of

Project Management Life Cycle” by Brown (2024). Retrieved from:

https://www.invensislearning.com/blog/5-phases-project-management-lifecycle

Following this, we would have the planning phase of the project, where it would be

broken down into smaller activities, which would be detailed as much as possible, and this would

allow a clear picture of the different steps and approaches to be undertaken to achieve the
32
objectives established. Each of these activities would have different resource requirements, and

in this phase, plans are made to procure the said resources. During this phase, the project

manager creates a project plan that details the tasks, dependencies, activities, and dates.

Additionally, the project manager coordinates the preparation of a budget that predicts costs for

workforce hours, equipment, and supplies. This budget serves as a tool to regulate and oversee

expenditures. The team selection for the different activities is carried out by the project manager

at this phase, along with the different stakeholders. Identifying the stakeholders in the project is

critical as it allows us to understand the needs and role that each individual would play and the

power or influence they hold that could influence the success of the project (Missonier &

Loufrani-Fedida, 2014). This is because, in a project, there could be different people with

different levels of interest in the project, which could either provide assistance or act as a

hindrance. Thus, for the project's success, robust stakeholder management is critical.

During the planning phase, risk management is undertaken, which is focused on

identifying both external and internal risks that could impact the project and taking the required

measures to mitigate or avoid the risk that could hurt the project and its success. Identifying the

risk is carried out by analyzing past projects, exploring the business structure, competition within

the industry, and external macro and micro factors and how they would influence the projects

(Jaafari, 2001; S. Q. Wang et al., 2004). The process of conducting the investigation, recognizing

any hazards, and developing a strategy to handle these scenarios are smoothly incorporated into

the entire project blueprint. A communication plan is also established, which discusses the

communication between the different stakeholders and covers how the project progress would be

reported and to whom it would be reported. The communication plan for getting people

33
committed and aware of their role in the project is critical, as many projects have failed due to

poor communication (Obikunle, 2002; Zulch, 2014).

The next step entails executing the project plan that was formulated in the previous

stages. During the implementation phases themselves, the monitoring and control phase would

begin. The progress is continuously monitored to assess, and any necessary changes are made to

the system based on the growing progress and encountered dangers. The project manager is also

responsible for collecting information from different teams, making reports based on the

information collected, and making necessary adjustments to the project to help present an

overview of the progress made to the stakeholders. Suppose modifications are made to the

project plan; these changes are often recorded along with why the changes were initiated (Project

Management Institute, 2017). It would help highlight what factors influenced change, which

could be helpful in future projects. In the last stage, known as closure, all project objectives are

met, the project is turned over to the next party, contracts are ended, and final communications

are finished.

Project Success

According to Bodicha (2015) one standard definition provided for project success is, if

the project is finished on schedule, meets the customer's expectations in terms of scope and

satisfaction, and stays within the allocated budget, then it has a good chance of being successful.

The triple constraint of time, 1scope, and cost are considered the iron triangle, and over the

years, it has been highlighted that limiting the success of these factors will not be effective

(Bodicha, 2015). This is based on the fact that there would be some projects that would deliver

all the functionalities and have met customer satisfaction, but the time and budget have been

overrun. Will the above scenario be preferred over a scenario where the budget and time

34
constraints were met while the project was not satisfactory for the customer? Studies that focused

on software project success and failure found that in specific scenarios, there were projects

canceled but were considered a success by developers (Linberg, 1999; Procaccino et al., 2005). It

is found that the lessons learned from the project could be used as guidelines for future projects.

It is to be noted that over the period with the product, the customer’s perception of the project’s

success could vary. It has been observed that when it comes to some project managers, there is

more focus on the wishes of the customers, be it based on their satisfaction or specific factors or

priorities that would influence the overall project success (Agarwal & Rathod, 2006). As a result,

predicting success or failure becomes a critical issue, as the variables that matter most for

determining variation and change depend on the perception of the project owners. There have

been multiple models like the Project Success Analysis Matrix (Demir, 2011) and others that

have been developed to improve the prediction. However, setting a universally accepted project

success criteria system is complex and impractical. Thus, the standard tool of cost, time, scope,

and customer satisfaction is often used to measure success. The abundance of data accessible in

the field makes it possible to make predictions, even though not all measurements are reliable.

35
Figure 4

Iron Triangle

Note. Showing the three critical success factors of Project management. Inspired by the Original

work “Iron Triangle” by Dhillon (2018). Retrieved from

https://medium.com/@harpreet.dhillon/iron-triangle-triple-constraints-of-project-management-

e818e631826c

Agile Project Management

The project management that occurs in sequence and is developed based on the needs of

the construction or manufacturing industry, where few changes would be required across the

different stages, is referred to as the waterfall methodology (AltexSoft, 2016). The waterfall

method is heavily focused on planning and specification development, which is said to take

around 40% of the total time and budget of the project. This method is most suitable when the

projects are well-defined. The waterfall approach requires a lot of planning, extensive

documentation, and tight control, especially in the development process. That said, due to the

lack of flexibility, the project model is not ideal for projects requiring software development

changes. Due to the various restrictions, the waterfall method would be slow, inflexible, and

36
costly for these projects. The method’s inability to adjust to the product and the changing market

requirements is responsible for resource wastage and the eventual failure of the said project.

Figure 5
Waterfall model (AltexSoft, 2016)

Note. Providing an overview of how the waterfall methodology worked in consequent steps.

Inspiration from “Waterfall Model” by AltexSoft (2016). Retrieved from

https://www.altexsoft.com/media/2016/04/Agile-Project-Management-Best-Practices-and-

Methodologies-Whitepaper.pdf

To achieve flexibility, the agile methodology was introduced, and it was reported that by

2015 around 95% of the organizations had shifted to the agile method (AltexSoft, 2016). Though

the agile philosophy has been in use since 1957 in IBM and Motorola, the approach gained

popularity in 2001 when around 17 professionals in software development decided to meet and

develop a new process to create an alternative project management methodology (AltexSoft,

2016). By utilizing their extensive knowledge of industry demands, they devised a strategy

marked by adaptability, nimbleness, and a collaborative focus. This approach, which was

considered to be innovative and groundbreaking, was first published in the book The Manifesto

for Agile Software Development (Beck et al., 2001). In contrast to more conventional
37
approaches, agile works in shorter iterations called sprints. Each of these cycles is considered a

small project, meaning it has a backlog and would consist of its phases but always fall within the

scope of work. At the end of the cycle, a product could be shipped, a product increment like a

software update. This would mean that in every iteration, newer features would be added to the

product, thus allowing for consideration of customer feedback and gradually growing the project.

With the features identified being validated at an early stage, the chances of failure are reduced.

The main aspects of the process are as follows (Beck et al., 2001):

• Flexibility: If the needs are different now, the scope of the project could shift.

• Breakdown of tasks: The project could consist of brief iterations.

• The importance of teamwork: everyone on the team would be laser-focused on

their own tasks and would work closely with everyone else.

• The work done inside the cycle is continuously reassessed to ensure the final

result is better. This is known as iterative improvements.

• Working together with clients: in such projects, the client is actively involved, has

the freedom to modify the requirements, and is open to the team's suggestions for

improvement.

Some of the benefits that are seen as a result of the changes that are bought about by

Agile methodology would include (Beck et al., 2001):

• Ability in managing the changing priorities

• Increase in the team productivity with the help of daily task allocation

• Improved project visibility as a result of a simple planning system.

The purpose of agile development was to develop an overall process for development

along with a management strategy that helps improve software development projects' overall
38
success (Lalsing, 2012; Nasir & Sahibuddin, 2011; Sheffield & Lemétayer, 2013). When we

compare the predecessors to agile, it can be seen that Agile is far more helpful. Based on the

Standish group report, it has been mentioned that 39% of agile projects were successful, and

52% of them were completed but had a few challenges (Standish Group, 2015). The following

most popular method is the waterfall method which reported a success rate of 11%, with around

60% of the projects being challenged (Standish Group, 2015). Another report around the same

topic that compared the different project management methodologies found Agile to provide

55% of success while 36% had faced some challenges and just 3% failed (Ambler, 2018).

Though Lean methodology was found to have a better success rate at 68%, the number of

failures in the method, too, was high at 11%. Traditional methodology only saw 29% of the

projects successful, and 67% were challenged, showing the capability that agile provides to the

users.

39
Figure 6

Project Success based on Methodology

Succes Rate of IT Delivery Teams

Traditional/Ad Hoc 29% 67% 5%

Iterative 36% 42% 21%

Agile 55% 36% 3%

Continuous Deliver/Lean 68% 21% 11%

All teams 52% 40% 8%

0% 20% 40% 60% 80% 100% 120%

Succesfull Challenged Failed

Note: Showing the effectiveness of different methodologies in project success. Inspired from
“Perceived IT Projects success rates” by Ambysoft (2018). Retrieved from
https://ambysoft.com/surveys/success2018.html

The Agile methodology is distinguished by its focus on incremental and iterative

processes, as well as the accompanying management practices. These techniques facilitate the

achievement of higher quality and timely delivery by promoting ongoing improvement

throughout the whole duration of the project (Drury-Grogan, 2014; R.Raval & Rathod, 2014).

Agile encompasses a variety of methodologies, each based on distinct iterative ideas, which are

all crucial for achieving success in projects. Researchers have confirmed that these agile

principles make a substantial contribution to the overall success of projects that adhere to agile

methodology (R.Raval & Rathod, 2014). Due to the need for the projects to be flexible and

adaptable to the changing requirements and circumstances, information in the form of data at

different phases plays a critical role when it comes to decision-making. In agile methods like

40
scrum, the idea is put into effect with the help of short iterative feedback cycles that would rely

on the capabilities of the team members and the use of data that has been gathered to ensure

evidence-based decision-making. The scrum guide that highlights the need to gain knowledge

via practical experience and make decisions based on data is developed on the empirical process

control theory. This approach provides a concrete grasp of aspects that are already known

(Schwaber & Sutherland, 2011). Machine learning is clearly shown to be critical and effective in

their ability to make decisions, which is a critical part of the project management practice and

thus important to this study.

Data-driven decision making

Referring to extensive analytics and the use of data in providing answers to a specific

question and making decisions based on the analyzed data or information is referred to as data-

driven decision-making. The decision-making process is transformed under this technique, which

in the past relied on an individual's history of experience, intuition, and gut feelings, will

supposedly be superseded by this instrument. As a result of the lack of techniques available for

analysis. One of the main ideas of the concept is that the decision, along with efficacy, can be

deduced with the help of critical data sets. For example, when we consider an employee that is

working in marketing who is responsible for the development and selection of advertisements

that are to be shown on the website, visitors of the company could use their experience in

working and choose the advertisement based on their understanding of the different impacts the

different type of market campaigns has. When the DDDM is used, the employees can analyze

how the website users have interacted with previous advertisements, see what type of

advertisements are more attractive to the consumer, and develop and push an advertisement

41
based on this. While these methods are not considered to be exclusive, they are often found to be

used parallel, which helps in improving the outcome (Provost & Fawcett, 2013).

There have been increasing works focusing on decision support in software projects

based on data-driven systems. This leads to the view that data is the guiding force behind the

direction in which the project would be headed. This practice is considered a deviation from

relying exclusively on the intuition and subjective experiences of team members, which are

frequently regarded as unreliable and influenced by personal biases. That said human-centered

skills like imagination, problem-solving prowess, and empathy for stakeholders and end users—

who are ultimately customers—are essential for software initiatives. As a result, being well-

informed is crucial, and when making decisions, it's even more important to examine the data

and offer options for how others may perceive it. This view aims to ensure that data just provides

one part of the equation and could be hiding information that is not provided within the data and

that is missing that an individual with experience might recognize. Hence, having a balance is

highly critical.

In agile, decision-making is changed from the traditional system. This methodology helps

shift from the traditional command and control attitude towards decision-making based on data

and shared control, which would consider all stakeholders to be part of the project. This allows

the possibility of having smaller teams that have the required links to communicate with the

customers. Considering it to be more naturalistic, it is also mentioned that the decision-making

approach is not rational. It is to be noted that rational decision-making is based on the

assumptions that there are a set of solution possibilities and that there is also a probability of

known outcomes. Conscious analytical examination and the application of norms that depend on

context are given less weight in naturalistic decision-making, which is largely influenced by

42
situational behavior. Because of the significance of awareness and the project environment, agile

principles are in harmony with the natural process (Matthies & Hesse, 2019).

To achieve success, various types or forms of data are often used. These would include

(Matthies & Hesse, 2019):

• Business Data: companies often collect detailed information from customers,

suppliers, partners, and competitors. These data could be used to identify patterns and

infer relationships critical to project development (Brynjolfsson et al., 2011). The

user interface interaction and keyword searches would reveal insights into consumer

behavior without needing to run different surveys separately, which can be costly.

These types of data are seen in all companies, and increased digitalization has started

playing a critical role in business decision-making. There is also data collected from

consumers based on their experience with a particular product and looking to

understand their perception regarding the brand or product and what changes would

be required in the next iteration. This is not limited to software companies but can be

seen in other areas.

• Project Data: the projects like software development are said to have a plethora of

data that could provide valuable information. Previous project data, which would

include codes or project management challenges faced and how these issues were

tackled along with other reasons for a project failure, could be critical data from

projects that could be used in developing alternative solutions or strategies.

These are the two primary data sources not limited to the software industry but are

applicable in other industries. In Agile teams, self-management and self-organization concepts

often underpin the decision-making issue (Moe et al., 2012). Within a self-organizing Agile

43
team, the team members not only engage in the execution of diverse tasks but are also inherently

entrusted with the responsibility of independently overseeing and tracking their individual

contributions to the collaborative effort. In the same way, it is noted that the responsibility

regarding the decision-making too would be required to be distributed. Here in the methodology,

the power or decision-making capability is reduced for the project managers, and all stakeholders

have a say in the decision process. The concept of self-organization is said to impact the team’s

effectiveness as the authority for the decision-making would be directed to the lower levels of

operation, which would help improve the speed with which the problems or issues would be

addressed (Zannier et al., 2007). Despite the availability of data to improve decision-making,

studies suggest that Agile team members often rely heavily on their experiential expertise. In

order to foster flexibility, the Agile framework creates a setting where teams may proactively

create decisions that are both practical and strategic (Drury et al., 2012).

Machine Learning

As previously mentioned, we explored the importance of data, emphasizing the current

period in which we live, where everything is supposedly connected to a data source, capturing

and using every event digitally (Cao, 2017; Sarker et al., 2021). The data obtained from these

systems can be categorized as structured, unstructured, or semi-structured. It is crucial to extract

significant insights from this data in order to construct intelligent applications in relevant fields.

By employing data management tools and methodologies, we may efficiently extract and

effectively apply valuable insights. The importance of the tool in data analysis and computing

has increased as the algorithms allow programs to learn and develop independently based on

experience without requiring explicit programming (Sarker et al., 2021; Sarker, Kayes et al.,

2020). It is generally considered a critical technology in the fourth industrial revolution. This

44
new way of thinking places an emphasis on intelligent technology, such as automation driven by

cognitive algorithms, which incorporate complex data processing into traditional manufacturing

techniques. When it comes to making practical applications and doing insightful analyses of

data, machine learning is crucial. The built-in algorithms that make up this development fall into

four main categories: supervised, semi-supervised, unsupervised, and reinforcement learning-

oriented (M. Mohammed et al., 2016; Sarker, 2021).

Figure 7

ML Techniques

Note. Shows the different types of Machine learning techniques under different approaches.

Inspired by “Various Type of Machine Learning Techniques:” by Sarker (2021). Retrieved from

https://doi.org/10.1007/s42979-021-00592-x

Supervised learning is typically said to be that task for machine learning that would learn

the functions that would map an input to an output. In order to infer a function, this method uses

labeled training data and relies on identified sample components. Applying a goal-oriented

approach, supervised learning is used when specific objectives are defined and achieved using
45
predefined input datasets. Supervised tasks often involve data categorization and regression,

where the acquired data is used to predict the class label or sentiment of a tweet, for instance

(Sarker, 2021). Unsupervised learning is said to provide the ability to analyze unlabeled datasets

without needing them to be programmed or, in short, without interference, making it a data-

driven process (Han et al., 2012). This technique is said to be critical and helpful when it comes

to extracting generative features, identifying trends that are found to be meaningful and

structures, exploratory purposes, and understanding the role played in results groupings. One of

the most common forms of unsupervised ML techniques would be clustering, feature learning,

and anomaly detection.

By making use of both labeled and unlabeled data, semi-supervised learning integrates

the approaches of complete supervision and no supervision (Han et al., 2012). When there is a

surplus of unlabeled data but a dearth of labeled data, semi-supervised learning becomes

extremely useful in many real-world contexts. Improving prediction outcomes by using just

labeled data from the model is the primary goal of semi-supervised learning. This technology can

be applied in fields such as fraud detection and machine translation. Software agents or

computers can improve their efficiency using reinforcement learning, a machine learning method

that lets them study and assess the best action to take in a given scenario on their own (Kaelbling

et al., 1996). This learning methodology is based on a system of rewards and penalties, with the

goal of extracting valuable information from interactions with the environment in order to guide

actions that optimize rewards and minimize dangers. This approach is commonly used to train AI

models and has been found to be a powerful tool for improving operational efficiency in critical

and advanced operational systems, and one example of this can be seen in robotics. However, it

46
is also not considered ideal for solving specific straightforward or essential problems (Sarker,

2021).

One of the significant challenges that are seen in data science is the high dimensional

data processing. Therefore, it is necessary to utilize dimensionality reduction, which is an

essential unsupervised learning method renowned for improving interpretations, decreasing

computational expenses, and mitigating overfitting and redundancy issues that can arise in

specific models. Feature selection and extraction are tools that could be used in dimensional

reduction. When we compare the two processes, it is important to note that when it comes to

feature selection, it will retain a subset of the originally defined features, while the extraction

method is focused largely on developing a new set (Sarker, Abushark, & Khan, 2020; Sarker,

Abushark, et al., 2020). With the capabilities and techniques, machine learning has been

identified to be capable of multiple applications, which include:

• Predictive analysis and the capability of intelligent decision-making: The most critical

use of machine learning is to improve the decision-making process, which is often helped

by information analyzed and gathered from the vast amount of data in the form of trends

and other statistical measures (Mahdavinejad et al., 2018). It is based on the capability of

capturing and exploiting the relationships that are found between the exploratory and

predicted variables based on previous events that would help predict the outcome that

would be unknown. Machine learning is extremely important in jobs such as identifying

suspects after a crime or detecting fraud. When it comes to retail, algorithms are crucial

because they help stores better understand customer tastes and habits, which in turn

improves marketing and inventory management. Algorithms like artificial neural

networks and decision trees are said to be used for this purpose. Due to the accuracy of

47
 the prediction, the data and information are said to help improve decision-making

irrespective of the industry or its use in any of these sectors (Sarker et al., 2019; Witten et

al., 2016).

• Cybersecurity: This is also based on the capability of learning and analyzing data and

predicting behavior with accuracy, allowing for better detection of malware or identifying

traffic patterns that would help secure the data (Sarker, 2021).

• Smart city applications

• Traffic prediction and transportation

• Healthcare

• E-commerce and product recommendation

• Sentiment analysis

Although there is a rising tide of support for machine learning, it is critical to recognize

that there are a number of obstacles to its widespread use. It was noted that the field would face

challenges in data privacy, sharing, and accessibility (Jordan & Mitchell, 2015). These would be

highly data-specific challenges, but there is a need to understand the technical and managerial

challenges that the adoption faces (I. Lee & Shin, 2020). The ethical dilemma ranks high among

the obstacles. In spite of machine learning's widespread praise, there are still major differences in

how algorithms are designed and implemented, as well as a complex understanding of the ethical

implications, which are seen as having a significant impact on people, companies, and society as

a whole (Mittelstadt et al., 2016). Ethical considerations emerge at every stage of application

development. In order to gather and train data in a way that complies with privacy and protection

regulations, businesses must establish procedures. During the development phase, algorithms

have the potential to inadvertently acquire biases from the training data, underscoring the
48
significance of resolving bias concerns. Modern firms encounter difficulties arising from the

complex operating landscape, intense rivalry, demands from stakeholders, and obligations

towards society. It is essential to highlight the significance of ensuring that algorithms adhere to

ethical principles similar to those followed by human workers. This is necessary to ensure

consistency in the organization's ethical standards. There have been increasing standards and

protocols developed for the ethical use of machine learning. Managers would need to develop

corporate ethical standards and practices that would be used to recognize and minimize the

adverse effects of the biases within training data sets and ensure that the processes comply with

the ethical standards during the application development (I. Lee & Shin, 2020).

The scarcity of qualified engineers specializing in machine learning is another key

obstacle to its widespread us (I. Lee & Shin, 2020). Just in three years, it was observed that the

demand for AI engineers has increased by over 119%. Though the demand for engineers has

increased, there has been a significant drop or separation in the skill gap. In an effort to close the

skills gap and produce more engineers, universities have been concentrating on enhancing

existing programs and creating new ones. However, the industry's standards or requirements are

limited, and the demand for engineers is increasing. Another obstacle is data quality, which is the

data's appropriateness for use in machine learning. Research highlights that the effective

utilization of machine learning relies on the presence of data that is of superior quality (Katz et

al., 2014). Data quality is essential not only during the training process but also during

operational use. Ensuring the integrity of data is crucial for building trust in the program and

the conclusions it generates (I. Lee, 2017). Data quality is critical in deep learning systems. With

the increasingly unstructured nature of data and its numerous sources, the overall quality of data

tends to decline. The advent of unstructured data poses challenges for traditional data-

49
management methods despite the fact that structured data is crucial for machine learning. For

machine learning systems to work, unstructured input must be quickly and efficiently detected

and processed. Establishing dedicated processes that prioritize quality control is crucial.

Incorporating metrics for data collection and analysis, training set inconsistency detection and

removal, and cost-benefit analysis of quality assurance into these procedures is essential (I. Lee

& Shin, 2020).

A major roadblock to machine learning's widespread application is the cost-benefit

conundrum. Machine learning initiatives are anticipated to see a doubling in use from 2017 to

2018, according to research. However, the majority of organizations have only implemented a

small number of deployments or pilot efforts (I. Lee & Shin, 2020). Although machine learning

holds the potential for greater advantages, demonstrating its worth to stakeholders remains

challenging due to the time lag between investment and payoff. It is critical for managers to

understand that machine learning is not a solution that is suitable for all, and there is a need to

develop unique solutions or algorithms based on the situation. They need to carefully choose

projects and be able to justify spending. Conventional investment methods frequently fail to

accurately assess the true worth of machine learning projects. Choosing a real-option strategy not

only overcomes this constraint but also enables managers to optimize project planning

efficiently.

The above are some of the system’s challenges and are not specific to project

management but machine learning in general. These challenges have a role and influence and

thus are critical in understanding and considering adoption.

50
 AI, ML, and Project Management

Munir (2019) highlights the multiple tools available or provided to project managers

using AI. Some of these tools would include:

Chatbots: Chatbots are considered one of the main areas with many studies and focus

(Danysz et al., 2019). For example, it is said that Redbooth and Cisco developed a chatbot

platform that allowed team members to ask different things, like what is urgent, allowing for

better communication.

Strategies: Strategy is a tool that provides functions like slack to project teams and is

considered a virtual assistant-type tool that could find an increasing presence in the coming

years.

Zivebox: These are more complex software systems that provide the backend system's

AI functionality. It is believed that Zivebox, a digital workforce tool, uses AI technology to track

and determine how long it would take to complete a particular task (Flick, 2015). This also helps

to measure the productivity of all the team members and is said to help in sorting through the

communication databases.

Rescoper: This is another tool that assists in handling tedious parts of project

management to ensure that the team stays focused. This platform would tell the AI system what

to do and helps in the scheduling of the tasks and distribution of workload. Notifications are also

sent out in the event that the system detects a potential threat to the project's success or detects

that it is beyond the allocated time or money.

Clickup: this is another tool that uses algorithms to predict which team member would be

ideal for a particular task and would also help assign and manage the task load between the

different members.

51
 Polydone: this is another tool being developed that can automate the budget and the time

accurately required for completion.

The above tools highlight the various capabilities of AI-based tools and how they could

influence the project management process. Thanks to AI, project managers can now see which

projects are in a critical state and what has to be done right now to ensure their success (Prifti,

2022). The central aspect of the advantage provided would be the reduction in the response time

towards the project issues that are identified and are found to be outside the acceptable limits. It

has been mentioned that extra work is often done by undertaking overtime regardless of the

changing technological requirement, resource constraints, and skill requirements without AI

being used. By enhancing organizational value and identifying efficient resource management

techniques, AI may help facilitate the evolution of project portfolios, which in turn aims to raise

project profitability. Notably, this specific goal is served by purpose-built algorithms that are

utilized by a large amount of the program. That said, project scheduling and optimization

practices are said to be done manually, and some significant limitations or errors are caused by

the manual method due to trial-and-error strategies. It is believed that AI can aid in schedule

optimization by cataloging all of an organization's tasks and comparing them to a small subset of

those projects. This means more information is used to make assumptions (Prifti, 2022).

According to current theories in project management, it is possible to keep tabs on all the

assumptions and limitations as the project progresses. Because CEOs don't always know who

can step in to help with a project, AI is essential for solving problems and offering support in

these areas (Kerzner, 2014). Crisis dashboards are commonly used by enterprises to track various

project metrics. By analyzing whether metrics are not within an acceptable range, necessary

actions may be taken to remedy the issues. Such monitoring improves reaction time to some of

52
the aforementioned circumstances that would be regarded beyond the tolerance thresholds, as

previously stated, in scheduling and solution capacities. Management is said not to identify how

much additional work would need to be done or would be in the queue without overworking the

labor force. Often it is said that the projects are done within the queue, and there is less focus on

the technology required, skill levels required, and availability of the resources. AI-based systems

like ML allow for developing project portfolios that have the best chances of increasing the

business value, which would also identify the most effective practices suited for resource

management. Though many Software’s are available for this purpose, the activity is still mainly

undertaken manually and often based on the trial and error method as specified before (Kerzner,

2014).

AI is said to provide project managers with many benefits, which include:

• The use of AI is in place to automate the activities for the project managers as well as

provide feedback and alerts that help manage the workflow and ensure repeatable

activities are better handled. The AI future is said to include the support provided for

the workforce through the undertaking of some complicated approach to the

workflows for the products that are often spotted to be wasting time and also in

performance evaluation (Schwarz et al., 2015). It is also mentioned that quantification

allows for improved and more accessible analysis that helps support decision-making.

• Often, when gathered manually, it is reported that the data would be patchy, meaning

that some employees would give minute details regarding their tasks while others are

more diligent. AI could help provide more assistance based on the data given and

simplify the process, allowing the participants to provide more precise data,

increasing the accuracy of the information and thus improving decision-making.

53
• AI allows project managers to take on more complicated tasks that would not have

been possible before. The project managers would be able to get alerts regarding due

dates, which would permit the project managers to develop teams for particular

projects and determine the tasks to be assigned to each employee daily. The ability of

software is used to collate the data that allows anticipation. Multiple programs could

be used in tandem or separate which would ensure the errors would be reduced or

improve the overall experience and quality. Thus, AI allows for new insights and

strategies that were impossible with existing techniques.

• AI is said to eliminate the emotional and individual biases in the information

presented and rely effectively on data or critical information without any bias.

• Project Managers are often found to be bogged down by the numbers and would often

ignore the human aspect of management. Managers with the use of AI could leave the

data analysis to AI and focus on the management of people and other strategic aspects

of project management. AI is said to be faster than humans, with IBM Watson, an AI

model able to analyze over 22 million pages of text in just three seconds (Wu et al.,

2014). There is no doubt when it comes to the benefits of the system and the time that

project managers save from automation of the activities that could be used in other

areas, which could improve management. In addition, it is mentioned that when the

manager entirely does the analysis, they are so engrossed in the data that they often

fail to consider the emotional intelligence factor. There are projects that, in terms of

numbers, could look attractive but fail due to issues with the workforce not meshing

well with each other or some other aspect of the project which was not conveyed by

54
 data (Vanhoucke et al., 2016). AI allows the focus of project managers to shift to

these aspects while they do the number crunching faster.

• Considering that the project managers would have access to more information and

more time saved due to not doing the analysis, it provides them with more time to

focus on developing unique solutions. The information that highlights data that would

previously not have been analyzed allows for identifying new areas and providing the

freedom to be creative (Munir, 2019). The measures of data would allow them to

make changes as required making them more effective in developing innovative

solutions. AI also has improved predictability; thus, these solutions could be run in

simulations to see which of these newer or creative solutions would be better and how

they could impact the project.

ML Technologies in Project Management

The above is how AI could help project managers. We now look at how AI technologies,

including ML, influence project management. We have already established that AI helps

minimize errors in projects. This was especially seen in software development projects where

there would be a variety of defects that can be seen at different stages and is also a measure of

project quality. The use of AI in project management supposedly improves decision-making by

allowing for a better understanding of potential consequences. The technology identifies

relations and trends within data and removes unnecessary information allowing the management

to focus just on specific data that is critical (Belharet et al., 2020). One argument in favor of AI's

centrality to project success is the tool used for control and monitoring in project management.

Some of these tools would include neural networks and Bayesian models. In a study that focused

on the capability of such tools, it was found that the AI tools were more accurate compared to the

55
traditional tools even though they are said to remain complementary as their primary purpose at

present has been just monitoring and control (Magaña Martínez & Fernandez-Rodriguez, 2015).

Studies like these have recognized that the fusing of existing base tools has become a trend to

tackle the models’ weaknesses and improve their strengths.

This would mean that there is still a lack of coordination process that is seen between the

different tools that would help in project management. According to another survey, most

managers are affected by scheduling methods and timeframes, which are the major challenges

affecting project management. The management employs inefficient methods of monitoring and

control in an effort to finish the project (Prasad & Vijaya Saradhi, 2019). The study emphasizes

that managers may improve their performance with the use of task-specific tools based on ML

technologies. The following are examples of such tools: adaptive boosting neural networks,

which are used for project success prediction; genetic algorithms, which are employed for

identifying essential routes in a project and the corresponding success requirements; and fuzzy

cognitive maps and Bayesian models. Tools like the Red-Amber-Green (RAG) status evaluation

algorithm making use of Natural language Processing and Genetic algorithms, could be used for

monitoring and tracking the progress of the projects but also for resource allocations.

In the construction industry, one specific tool used in project improvement has been BIM

which has been in use since the early 200s and takes into consideration the structural and

functional aspects of the building, which are then built into the 3D models (Irizarry et al., 2013;

Khan et al., 2021; Koseoglu & Nurtan-Gunes, 2018). This would allow for the virtual simulation

of the project before it is executed and provide an analysis of the different impacts that the

various factors would have on the project in its different phases. This is considered a tool that has

helped improve flexibility and better resource allocation that is said to empower the people in

56
creating the proper infrastructure and would help save time and costs. This is also said to help

see the real-time progress of the projects being carried out and analyze where the mistakes could

happen and thus could also be considered a monitoring tool. The tool is said to streamline tasks,

improve efficiency, reduce the scope for losses and improve the collaboration across the different

teams and channels of communication, which helps the construction sector reduce wastage

considerably and become more sustainable (al Hattab, 2021; A. B. Mohammed, 2022).

Above is already a software-based system that uses data analysis. There has been an

increasing push for AI integration, which is said to help reduce the risks associated with human

errors and thus improve the overall decision-making quality. ML is said to help improve the

overall safety in an industry considered to be accident-prone and mitigate the different risks. AI

is capable of detecting discrepancies and warning people accordingly, allowing for the

development of contingency plans and evacuation strategies (Mohan & Varghese, 2008; Tran et

al., 2021). In addition to this, ML could be used to analyze the data and develop a building

design that is more efficient and acceptable. Once the required information is provided , ML

systems can create accurate floor plans and models, which can also be integrated without

needing disruption (Su et al., 2021; Ugwudike, 2022). These integrated plans would then be

analyzed to identify issues arising from different projects.

Drivers, barriers, and benefits of ML in Project Management

The importance of ML and AI starts with the increased limitations or challenges faced

within the Agile system. There is a lack of support that is available in several areas of Agile

management, and these include (Dam et al., 2019):

Identifying backlog items: the items within the product backlog could be captured from

different sources. When it comes to agile, due to a large amount of heterogeneous data, it

57
becomes a complex task. In the system, for each newly developed backlog item, there would be

an inter-dependency that needs to be considered. The concern is that there would be 100

backlogs in traditional models in general, which would mean a lot of additional

interdependencies that have to be defined.

Refining backlog items: Some items often found in the backlog would be significant,

meaning there is a need sometimes to fit them within a single sprint. This would mean there are

need to refine the items to shrink the oversized items to smaller sizes. Teams frequently face the

challenge of relying on their intuition and expertise to accomplish the same results, despite the

fact that there have been many suggestions for improvement standards over time.

Sprint planning: another critical part of agile management that is challenging is sprint

planning as multiple complex factors are needed to be considered like a priority, business value

to the customers, and many others, including team member availability and capability. Sprint

planning requires a higher in-depth understanding of them and also their experience. This is

challenging to do manually, requires many man-hours, and is highly complex.

Proactively monitoring the progress and managing risks: as the sprint begins, it is critical

to monitor and explore the risks that arise with the progress made. Using high-level instructions

and subjective assessments is the primary focus of the present system. The ambiguity,

changeability, and interdependencies of the initiatives make it harder and harder to foresee

potential dangers. This would mean that there is currently a gap between the insights available

and actionable information which would help in providing concrete measures that help deal with

the said risks.

58
 The research highlights that AI, which includes ML, can tackle all these issues and use

multiple tools and capabilities to provide the solution. A proposed architecture for the above

challenges has been represented below:

Figure 8

AI-Controlled Agile Project

Note: Showing how AI and its elements play a role in controlling the Agile Projects. Inspired by

“The architecture of an AI-powered agile project management assistant” by Dam et al. (2019).

Retrieved from 10.1109/ICSE-NIER.2019.00019.

As we can see above, multiple tools are being used, including machine learning or deep

learning processes. The system is said to help tackle and eliminate all the issues associated or the

challenges mentioned before (Dam et al., 2019). Each of the different AI-based systems helps
59
automate the processes that require data. The information collected and analyzed is converted

and provided as actionable information, making it easier for those involved in making the final

decisions. Even risk monitoring and tracking provide far more information and regular updates

with more detailed information compared to traditional systems. Thus, the system’s importance

is focused on its capability to simplify the overall process and make the decision-making process

more effective.

Surveys on AI and its perceived usefulness have shown that AI would impact all project

management processes. Three main areas of implementation are seen, or the use of AI. These

include process automation, augmented analytics, and chatbot assistance. These three

applications are said to find a place in all project management processes (IPMA, 2020). In

process automation, it was found that time and quality improvements or change are where AI has

the most significant potential and could help project management. Augmented analytics is most

beneficial or impactful in understanding the risk and opportunities while helping in better

planning and control. Reports over the years based on surveys have suggested multiple drivers

for adopting AI. The increased available experience, the availability of AI-driven systems,

increased stakeholders’ demand for value and innovation, pressure to deliver and improve

overall efficiency in managers, and increased adoption of agile practices across multiple

industries are critical factors that are said to have been the drivers behind the adoption of AI.

Barriers to Adoption

Prior to delving into the impediments, it's imperative to grasp the catalysts propelling the

integration of AI into project management. It has been discerned that the primary driver behind

AI adoption is the imperative for heightened productivity. This is asserted to liberate project

managers, enabling them to focus on more pivotal decision-making endeavors (Bodea et al.,

60
2020). Another key driver has been attributed to the growth in data that would help in improving

decision-making, but it has become increasingly complex for manual techniques to be useful in

its analysis. As companies have become increasingly aware of the importance with regards to the

adoption of Big Data and its use in gaining a competitive edge, there has been an increasing push

for AI and ML practices that are known to provide or aid in analysis (Bodea et al., 2020; Cubric,

2020). The growth and evolution of AI and ML technology that has evolved into capable systems

that are flexible, capable, and more accessible has also pushed the adoption of AI systems. Cost -

effectiveness is also playing a critical role in the acceptance and adoption of the technology

(Cubric, 2020). With improvements in technology, the cost of computing and data storage has

decreased significantly, making AI solutions more affordable for organizations of all sizes. The

reduction in resources and increased efficiency also mean saving for the organization, which

further highlights the cost-effectiveness of the technology. The fast growth in the use of AI and

ML across sectors and applications is largely attributable to these factors (Bodea et al., 2020;

Cubric, 2020).

Though the popularity of AI and ML systems has been growing significantly, multiple

barriers that deter adoption include the management or stakeholders' limited understanding of AI

technologies. There is also the issue of understanding what best AI applications are available and

which is ideal for the company. The concerns related to the data and how the data would be used

are also said to be a significant concern for the companies as they are worried about hacking or

cyber-attacks that compromise data and leak client information or confidential business

information and thus try to avoid the adoption of technology where data needs to be collected

and stored.

61
Figure 9

Barriers to adoption

Note. Providing an understanding as to which factors the practitioners feel are the biggest

barriers to adoption of AI. Inspired by “The top 5 most important barriers in using AI tools in

PM today” by IPMA (2021). Retrieved from

https://www.ipma.world/assets/IPMA_PwC_AI_Impact_in_PM_-_the_Survey_Report.pdf

That said, most industry leads or PM professionals positively regard the adoption of AI

and believe that using AI would help improve productivity, decision-making, and overall

performance. These three are considered the primary benefit. However, there are other benefits

they feel will be gained, like improved cost management and resource utilization, flexibility in

project management and improved communication, and increased compliance (IPMA, 2020).

This perception is also shared in many other studies where interviews were undertaken. Based on

the responses, it was seen that the AI application would improve the data quality and integrity,

which would help improve the overall speed of the process, and it is reliable and valid for multi-

project environments (el Khatib & al Falasi, 2021).

62
 A study exploring barriers to adopting AI and ML in radiology practices found that the

respondents reported being unaware of AI and ML practices and their use in Radiology practices.

Both doctors and patients had a hard time understanding and embracing the technology due to a

general lack of knowledge and confidence in it (Eltawil et al., 2023). The study also brought

attention to other variables that might hinder adoption, such as the lack of a systematic

implementation process and uncertainty about the benefits linked to the system and the influence

it would have on patient outcomes. They were also worried about how the advancement could

impact their professional autonomy and a general mistrust regarding human-led decision-

making. When organizations or hospitals asked why AI has not been adopted, the common

response was the lack of acceptance among radiologists (Barreiro-Ares et al., 2023; Codari et al.,

2019; Waymel et al., 2019). There is a general lack of trust when it comes to technology, and it

was found that clients also prefer the information to be presented by the radiologists themselves

and trust those reports more than AI generated alone (Lim et al., 2022).

There have been different studies that highlight the major barriers to AI and ML

adoption, and one such study that used a survey found that 40% of the respondents mentioned

the lack of skilled professionals as one of the major barriers when it comes to adopting the

technology (H20.ai, 2020). The other two big challenges identified in the study were accessing

and preparing the data for machine learning and the limited budget available for the companies

to invest in or develop these systems. To elaborate on the issue, it is mentioned that the lack of a

skilled workforce means there is a lower number of professionals capable of helping the

organization in its development, and the lack of skilled people also drives the salaries as the

demand far outstrips supply. Upskilling existing developers and data scientists is a possible

63
solution but would only partially alleviate the pain point, and establishing an equilibrium would

take time (H20.ai, 2020).

The second issue of data access and preparation is quite understandable, considering the

large amount of data that ML or AI systems would require for developing the required systems

which involves training and testing. The lack of data in certain fields like construction is said to

act as a bottleneck at the beginning of the machine learning system design face, which could

derail the project. The costs are a significant challenge as the companies must hire new talents

and procure new infrastructure and tools (H20.ai, 2020). Even if they outsource the development,

significant infrastructure changes would be required to ensure efficient use of ML and AI

products. While they can provide a competitive edge, which prompts the companies to adopt

them, the high price tag is still a concern, and the fear of failure and change in the organization

deters them from adopting the tool. The fear can be seen based on reports highlighting that just

15% of all ML projects are known to succeed as of 2021 (Panikkar et al., 2021). It is mentioned

that most companies are still stuck in the pilot stage and are struggling to apply ML more

broadly to take advantage of the most advanced forms (Panikkar et al., 2021). Only 36% of the

projects are known to move beyond the pilot stage, highlighting the reasons behind the fear

among the executives.

A plethora of obstacles were exposed when the Indian public distribution system's AI

integration was examined. Problems with the new AI system include widespread skepticism of

technology, complicated politics, limited understanding of AI systems, inadequate economic

policies and regulations, a lack of competence, difficulties with integration, questions of

accountability and responsibility, fears of excluding marginalized groups, resistance from

employees, uninspired consumers, and outdated IT systems (S. Kumar et al., 2021). Another

64
study that surveyed 11,248 SMEs with a questionnaire achieved a response rate of 283

questionnaires. This study identifies three main factors as barriers: a lack of competence,

implementation challenges, and data problems (Ulrich et al., 2021). Another survey that explored

37,003 companies with more than ten employees from select industries found that 90.4% of the

companies do not even consider using AI technologies. At 58.1%, the information and

communication sector is said to be leading the adoption rate, which is said to be due to the

proximity of technology. The main reason highlighted for this has been the lack of expertise,

costs involved in the adoption, incompatibilities in integration with legacy systems, and the

complex data requirements (Rudolf, 2023).

The difficulties were greatest for SMEs, according to a survey conducted in Germany. A

survey carried out in 516 companies found that the biggest hurdle is the cost of the system and its

implementation, along with the lack of skills and the management of the legal and technical risks

associated with the adoption of AI. Data protection, lack of access to external data, and security

issues are said to play a critical role in limiting adoption (Rudolf, 2023). Construction-based

studies have highlighted the issues with regard to the lack of a common solution model in

different applications, along with the challenges related to ethical data privacy and protection.

Another difference has been attributed to the shift in the need for prediction explanation that

needed to be done in the past to understand the underlying learning techniques better and help

make informed decisions. This study also highlights that the lack of in-house capability and the

costs involved in the development are the major challenges to adopting AI or ML (Akinosho et

al., 2020).

Integrating artificial intelligence (AI) inside organizations presents a range of internal and

external obstacles. Four primary internal challenges need to be addressed . Organizational

65
leadership needs to be willing to adopt artificial intelligence (AI) technology and devote the

requisite resources to facilitate its integration. In the absence of this pivotal stage, the entire

procedure may experience a halt before its initiation. Another issue is the lack of knowledge or

awareness among the employees. The knowledge gap hinders progress and leads to failure in

identifying the relevant use cases. Another issue is the organization and its infrastructure. This

would include both financial and technological considerations. Most businesses are said to find

themselves constrained by various financial resources, which makes it challenging to invest in AI

expertise. Lastly, the final internal barrier is the data-related hurdles, which are based on the

complexity involved in acquiring and maintaining high-quality data while safeguarding it against

any security breaches or misuses (Rudolf, 2023). Only one external factor is considered a barrier:

the availability of suitable AI software in the market. AI possesses intrinsic practicality in the

context of organizations, hence underscoring the significance of two crucial criteria. First and

foremost, the organization must possess or have access to the requisite skills to effectively

discover possible applications of AI. Furthermore, after identifying key use cases and

establishing a comprehensive data foundation, the organization must decide to either develop AI

software or procure such software from external suppliers (Rudolf, 2023). A study in the medical

field shows that this sector has advanced rapidly and has been able to develop new solutions for

the market, thus reducing the external barrier to a certain level (Bahl, 2022; Rudolf, 2023).

A lot of research has been done on how to use AI and industry 4.0 tools in project

management. However, the factors influencing the uptake of technology in project management

have been the subject of scant research (Güngör, 2019; Shang et al., 2023). A specific study

centered on project management pinpointed significant hurdles in the adoption of AI. The

foremost challenges identified include the substantial costs associated with AI implementation

66
and maintenance, as well as insufficient support from organizational employees across all

hierarchical levels. The study further contends that a scarcity of adequately trained personnel in

AI poses an additional impediment to widespread adoption. The study was done in the building

and environment industry. The studies in the field across different industries are very limited

(Shang et al., 2023). The specific challenges related to project management are less explored in

the literature. This study would help improve the existing literature and provide specific

challenges based on perception across different industries, enriching the literature and knowledge

to better understand the common challenges across sectors.

Synthesis of Research Findings

The study's findings shed light on the complex relationship between project management

and the successful completion of projects. Cost, time, and scope evaluations have traditionally

been at the center of success assessment, as may be determined from a thorough literature

analysis. But these standards are becoming more inadequate for judging the success or failure of

modern undertakings due to the rising complexity of such endeavors. Success in completing a

project is still critically important, even though the conversation around it is changing in the

scholarly literature. Here, success is defined as sticking to the project's scope, meeting all set

objectives, maintaining a high standard of quality, and dealing with any changes to the project's

limitations. A critical component in this setting is the flexibility to adjust measures and evaluate

performance in reaction to these ever-changing conditions. The complexity of the projects has

meant the need to integrate new practices or methodologies that would provide flexibility. In the

IT industry that required faster adoption and improvements between the different iterations,

which led to the evolution or development of agile methodology. Agile is said to be a set of

values and principles focused on improving collaboration, self-organizing, and adaptability to

67
change. The focus on technology is said to be to face the uncertainty, determine a solution and

try the solution, gain feedback and make the required adjustments based on the feedback. This

methodology was also seen in other industries after the improvements it achieved in the IT

industry. Even in the most supposedly complicated business for project management, the

building industry made use of the agile technique. The complexity of the methodology with

increased time and commitment requirements is a significant pitfall, while the predictability for

the different factors, like resource requirement, is also a significant challenge.

It is evident from Agile’s shortcomings that data and analytics have the potential to

contribute vital information that might enhance decision-making as a whole. It was found that

data-based decision-making allows for gaining more information and making a more informed

decision. Big data has played a significant role in different applications, and its capability allows

companies to minimize the risks involved. The development of data analytics has provided the

predictability capability required in agile and reduced the workload of project managers as AI or

ML systems can provide the outputs required for decision-making. One of the technologies that

have been developed to minimize human commitment and increase the ease of operation is AI

systems.

We also explored the development and techniques used in machine learning and the

application it could use. This was critical in understanding the capabilities. We also explored in

the literature the different challenges related to adoption—the lack of skilled engineers, data

quality issues, cost-benefit challenges, and ethical issues. These are the significant issues that act

as barriers to adoption in many applications and are also considered to be shared for project

management. We explored the same challenges in AI-based adoption and the various

applications in which AI-based technologies have been used in project management. It shows the

68
capabilities of AI in easing the workload and improving efficiency. The machine learning

processes that have also been adopted in project management were explored, but most of these

are based on the capabilities and possible benefits that could be gained. Even case studies

highlight the capabilities but do not address the limitations and challenges faced in the adoption

or discuss what barriers must be overcome to achieve the goals. The extensive literature review

provides a holistic picture of machine learning by covering all the available studies. A key focus

of the project's inquiry, it highlights a clear gap in the literature about adoption difficulties.

Conclusion

The chapter explores the project management phases and success factors and how data

has been used in decision-making over the years. These changes highlight the growing

importance of data and information and the need for sustainability and reducing risk increasing

the push of AI systems. AI systems respond to the need to analyze the growing amount of data.

ML systems' capability to identify trends and provide structured and accurate information allows

project managers to improve their decision-making. As seen from the literature, AI systems are

support tools for project managers and cannot be used for decision-making due to the system's

lack of emotional intelligence. The main task identified for AI has been monitoring and control.

The system can be used in different phases for activities with increased accuracy, but its adoption

is still relatively less. The capability of AI allows it to be used in different phases at different

activities, and specific activities have been highlighted in the literature that could be a small part

of its potential use.

69
 CHAPTER 3. METHODOLOGY

Introduction

Clarifying the methodology used to gather and analyze data for the study is the main

focus of this chapter. The study's methodology is thoroughly examined, including its processes,

participants, settings, examination of research questions, credibility, reliability, strategies for data

collecting, tools for data analysis, and ethical issues. Providing a thorough understanding of the

study approach is the main objective of this chapter. The information is organized into several

subsections that explain the research methodologies used to answer the study's research

questions in depth.

Design and Methodology

A research design is a strategy outlining the study's many components and the

researcher's intended methods of data collection and analysis, with an eye on adhering to all

applicable ethical norms and regulations. Researchers often divide their work into two broad

categories: qualitative and quantitative. This is dependent on the data collected for the study

(Kothari, 2004; R. Kumar, 2017). Quantitative research uses numerical data collected by

experiments or surveys, and numerical data analysis is undertaken provide the answers. On the

other hand, qualitative deals with non-numerical data, including texts collected through tools like

interviews or literature reviews and analyzed to provide the required answers. This study is

qualitative research. There are six standard qualitative research designs which are (Astalin, 2013;

Patten & Patten, 2018):

• Phenomenological studies: These studies focused on examining human experiences

through the descriptions provided by the people involved. Here, the focus is on

70
 expressing or interpreting the meaning behind each experience expressed by various

individuals.

• Ethnographic studies, which center on gathering and analyzing information on

specific cultural groups. Ethnography research is undertaken to show how the actions

in one world would be perceived from the point of view of another culture or would.

• Grounded Theory Studies: These investigations seek to develop a theory that is

grounded in the collected data through data collecting and analysis. It is believed that

this strategy will use a combination of inductive and deductive techniques to build

hypotheses. Data are often gathered in their naturalistic settings, and the tools often

used are observation and interviews. Data collection and analysis are said to occur

simultaneously. A process commonly referred to as constant comparison would be

used in which the data is often compared to the data that has already been gathered.

• Historical studies: these are studies focused on the identification, evaluation,

synthesis, and location of data linked to past events or studies. The research not only

focuses on discovering the events but also aims to relate these happenings to the

present and future.

• Case-Studies: this is said to be an in-brief exploration focused on a specific subject

that could fall under either the qualitative or quantitative method . These studies are

said to be effective in describing, evaluating, understanding, and comparing the

various aspects of the research problem.

Considering that we are going to be exploring the opinions and perception of how AI and

ML are used in project management and what they believe the benefit experiences of these

individuals is ideal, which makes it a generic qualitative research and is close to a

71
phenomenological study. It is also a robust tool in situations where holistic and in-depth research

is needed (Zainal, 2007). Conducting generic interview-based research emerges as an ideal

approach when seeking to explore participants' experiences with technology. This method offers

several advantages that contribute to the depth and richness of the gathered insights. There are

many benefits to this type of approach which includes:

• Extensive Analysis: Interviews allow people to open up about their complex and

nuanced opinions on technology. Researchers can learn a lot about the topic by

probing participants' thoughts and feelings.

• Getting a Feel for the Big Picture: The interview style permits an in-depth

investigation of the setting in which people use technology. Accurately assessing

their experiences and illuminating the external aspects impacting their relationships

requires this contextual expertise.

• Adaptability and Flexibility: Researchers can change their asking strategy depending

on participants' answers since interviews are naturally flexible. When conducting

interviews, it is crucial to be able to quickly adjust to new circumstances in order to

go more deeply into particular topics or uncover unexpected themes.

• Interviews capture the depth of participants' experiences and give rich qualitative data

that goes beyond numerical measurements. Understanding the subjective aspects of

technology use, including feelings, views, and goals, requires this wealth of data.

• Interviews encourage participants to feel empowered and to take control of their story

by letting them describe their experiences using their own words. By focusing on the

participants, we can ensure that the data we collect is genuine and accurate.

72
 • Interviews provide a great opportunity to delve into a variety of viewpoints held by a

sample of people. For differences in experiences that would be hard to see in

quantitative research, this qualitative approach works effectively.

The wide array of sources and tools for data collection potential that the approach

provides highlights why the case study approach is ideal. Data triangulation validity would also

be addressed but would provide a broader understanding of existing literature and perspectives of

experts working in the field.

Participants

Often in this type of research, it is critical to ensure that only highly qualified people

considered experts in the field are considered as participants. Expertise is considered a level of

proficiency that an individual possesses, which novices have (Chi, 2012). A person is regarded

an expert in their field if they have extensive experience, are well-respected by their colleagues,

make exceptionally sound decisions, demonstrate flawless performance, and are able to handle a

wide variety of unusual or difficult issues. One further definition of an expert is someone who

has amassed a wealth of knowledge and expertise in a certain area through years of dedicated

work (Chi, 2012). In this study, the focus is on ensuring that participants are Project

Management Institute (PMI) members with at least ten years of experience and reported having

experience using AI or ML technology when it comes to project management activities. These

individuals would be contacted and based on their willingness to participate would be shortlisted .

Regarding expert interviews, the number of participants required for such has rarely been

established (Baker & Edwards, 2012). It is believed that the number of participants should not be

too small that it would lead to data saturation, nor should it be too large, which would make an

in-depth analysis of the collected data difficult (Leech & Onwuegbuzie, 2005). Various studies

73
have been undertaken to explore the ideal number of participants and have shown different

ranges. Some studies believe between 12- 60 interviews, others believe it should be five to 25

participants, and others mention that the total number of participants should not exceed 15

(Saunders & Townsend, 2016). One study that focused on summarizing the total number of

participants for the research identified that anything ranging from four to twelve would be

sufficient when choosing from a homogenous population (Saunders, 2017). Considering that the

participants identified are similar and have similar traits, they can be considered homogenous,

and thus a 4-12 participant range would be ideal.

On top of that, we would be referencing previous research on machine learning's impact

on project management to shed light on its many applications and drawbacks. The studies would

be selected based on keywords, machine learning, AI or Deep learning, and Project

Management. Only studies that illustrate the use or elaborate on tools used for the purpose would

be selected, and the rest would not be shortlisted. The research would aim to identify at least ten

studies that would elaborate on the current use of machine learning.

Setting

The interview will use Zoom video, but record audio only. The interviews would be held

online after communicating with individuals, getting the required permissions, and getting the

information from the participants to participate in the study. Video conference allows them to

participate in a more relaxed environment. There would be a set of 12 questions designed to

elaborate on their experiences and perception regarding machine learning and its influence on

project management, which also explores their second-hand experiences, the challenges faced,

and what they believe is the future for the technology. The questions would be open-ended, and

the responses would be recorded online and later used. The recorded information was converted

74
to a text transcript, which would be used in the content analysis where themes are developed. We

make use of standard software for highlighting and developing themes.

Analysis of Research Questions

The interview works on the central research question, which focuses on the benefits and

barriers to adopting AI and ML technologies in Project management. The research question was

developed based on the need to explore the perception of experts in the field regarding the

perceived usefulness of these technologies that would aid businesses in improving their project

success and overall organization efficiency during the different project phases. Understanding

the barriers and possible solutions based on those who have experienced the technology and

successfully implemented it in their organization would help identify the pitfalls and provide a

better understanding of how to simplify the process and be more effective for all involved.

Finding a solution would be the key, as from the literature, we know the same barriers exist in

different industries across different applications. With the growing number of projects and the

increasingly dynamic nature of the environment, AI and ML systems are critical in improved

decision-making. These technologies would also provide a competitive edge or equal footing for

smaller and medium-sized companies if they can adopt the technology successfully without

wasting money.

75
Table 1

Questions and Framework Link

76
 In the interview, the questions were designed to understand their experience and personal

views based on these opinions regarding AI and ML separately, along with understanding the

benefits they feel. This allowed us to gauge if the findings are similar to what is in the literature

and if the companies have achieved the same benefit in project management. Then we asked

about the barriers and the challenges to adoption, which in literature had limited coverage when

project management was in focus. This would help enhance and consolidate the views across

industries on the most critical barriers to be tackled. The questions then moved to possible

solutions, which is critical in understanding how they, based on their experience, could help

tackle these issues, along with exploring their perception that AI and ML in project management

are viable. During the in-person interview, participants were briefed on the study and its

objectives. In the lead-up to getting the required consent form, the participants had made the

individual aware of the possible time required for the interview, and sessions were arranged

based on a time frame they were comfortable with. In the interview, the questions were asked in

the same sequence across studies and the audio was recorded as agreed with the participants. I

ensured no personal questions were asked, and the comments between questions were limited to

letting them express their views. Before the interview began, I did mention to the participants to

provide their complete view and be as detailed as possible so that other questions are not

required. Some questions were repetitive but covered different aspects and helped emphasize and

cover any areas that might have been missed. Once the interviews were completed, the audio file

was stored in a portable storage system, which was then used in the transcription program to

convert to the transcript. All the audio files are still stored and would be kept secure and would

be deleted after three years. Overall, the research based on literature helped better understand the

common view in literature, and the questions provided a real-life experience-based perception,

77
which allowed for a better understanding of the situation and added to the quality of information

presented in the research.

Expert Review Panel

Before the actual interview, to better understand if the research questions would provide

the required answers that would help in providing the required answers for the research

questions, we would undertake an expert review. Under the expert review, the f ocus would be to

request the opinions of experts in the field of study to review the questionnaire selected for the

interview purpose and make changes based on their expert opinions. In this study, the research

would request an expert opinion from two people. Both would be professor-level experts with

machine learning and project management knowledge. After being briefed on the project's

background and objectives, they would be asked to review the paper. Then the questionnaire

would be presented to them for review, along with the expectation or answers the researcher

hopes to achieve from each question. Changes would be made if required based on their

questionnaire review and suggestions.

Test Run

For the rest run, the research would undertake a mock interview that is aimed at checking

the quality of the questionnaire and its ability to provide answers from respondents that would

help in deriving the required answers for the research questions. The researcher would request a

mentor and one colleague to participate in the process. These individuals would be selected

based on their knowledge in the field and past experiences. They would be asked to respond to

the formulated questions. The responses would be recorded and analyzed to check if the

responses are satisfactory and provide the expected required answers. The respondents would

78
also be asked for feedback, and based on their opinions, changes would be made to the

questionnaire to fine tune it before the interview.

Credibility and Dependability

Some of the steps taken to ensure the credibility of the study and an accurate record of

the phenomena include (Shenton, 2004):

• Random sampling of individuals is done to negate any researcher bias regarding

participant selection. Apart from the specification of experience and the PMI

presence, there are no other conditions, and the participants are selected randomly

based on their own willingness to be a part of the research.

• Adopting research methods that have been well-established in qualitative

research.

• Understanding the industry culture and participants’ organization industry and

how they undertake and complete their projects.

• It is also critical to ensure that only willing participants are part of the study as

there are chances of data corruption when unwilling participants are a part. It is

also critical to ensure that people are open and honest about their opinions on the

topic.

• Peer scrutiny will also be considered where a person or expert in project

management who is not a part of the study but is a neutral observer will review

the finished dissertation, but will not see any data collected or be part of the

research project.

79
 Data Collection

The researcher posted an announcement on the LinkedIn profile public news feed.

LinkedIn users viewed the announcement on the public profile and then contacted the researcher

if they were eligible. They were then provided with the screening questions. Those who

responded and met the criteria was sent an informed consent form using DocuSign for their

electronic signature. After I received the e-signed informed consent I scheduled the interview.

All interviews were scheduled at least 24 hours after the volunteer had received the form.

The researcher recorded the interviews using the Zoom video application and use Otter

transcription service to transcribe the interview. The researcher recorded the transcriptions,

transferred them to a USB drive, encrypted them with a password, and then put the disk in a safe

at the researcher’s home. The informed consent forms was likewise kept on a separate USB drive

that is password-protected and kept in the same safe. The researcher agreed to keep the data

secure for seven years in accordance with Capella University's requirements. After this period,

the researcher will physically destroy both USB devices by hiring a professional shredding firm,

which is both secure and ecologically friendly. The study data is handled with utmost care and

secrecy over its entire lifespan, thanks to this thorough approach.

Data Analysis

Considering that the research undertaken is based on textual data, one way to find the

required answers within the textual data is through the use of thematic analysis. The research

analyzes multiple literature and interviews, all of which would have critical texts and those that

are not critical to the research. Identifying the critical elements is essential. From the plethora of

data, there would be some common points. Identifying these common discussion points would

80
help in finding solutions or parts of the research puzzle. Thus, the thematic analysis would be

ideal in this scenario.

For qualitative data analysis, we use thematic analysis software like NVIVO or

MAXQDA. The TAM framework would guide, regularize, and support the analysis and

identification of the emerging themes on the limitations and recommendations for future machine

learning in project management. The data collected would be analyzed to identify common

themes highlighting the benefits and use of machine learning and the limitations and changes that

need to be made. The data from the case studies would also undergo the same thematic analysis

to see their use. The themes developed that could be considered as benefits, and limitations

would be compared to those seen from interviews.

Once the data was collected, the research went through the thematic analysis process. In

order to gain a deeper understanding of the project managers' acceptance, usage, and perception

of the benefits and advantages of technology, we utilized an inductive technique that allows the

data to assist in establishing the themes. Our focus was on the technology acceptance model. A

seven-step process is used in the thematic analysis, which is integrated with the TAM model

(Joffe, 2011; Kiger & Varpio, 2020). The first of these steps is to familiarize ourselves with the

data. The data was initially in the audio format and listened to at least twice. The data was also

transcribed with the help of a professional that allowed us to get textual data that could be used

in the coding process. Reading and listening to the interviews helped the researcher familiarize

with the data and understand various distinct or common perspectives.

The second step is the coding for which NVIVO was used. Coding refers to highlighting

sections of the texts that would include phrases or sentences and then using specific codes with

shorthand labels that would describe the content. Within the text, when we talk about benefits in

81
the interview, there are AI benefits as well as ML benefits and specific benefits that are talked

about by each interviewer, which all form different codes. These codes and concepts might be

similar across different transcripts, and each of these, based on the context and specific codes, is

first created in the second stage. Here the focus is on being as thorough as possible and

highlighting all those points that would be interesting or relevant. Once this is completed, all the

data would be collated together in different groups based on their codes, allowing us to gain a

condensed overview and the common meanings that might recur within the data across the

interviews. Here in this coding process, we used the codes to be developed after each interview.

As other interviews were coded as available, they were compared to each other to identify

common themes that exist.

Figure 10

Seven Steps in Thematic Analysis

Note. Showing the process of thematic analysis and how it would be undertaken. Inspired by

“The Seven steps in Thematic Analysis” by Kadir et al. (2021). Retrieved from

10.5057/ijae.IJAE-D-20-00033.

82
 The third step would be searching for the themes, and the fourth would be developing the

themes. To do this, the codes are looked at, and some common patterns among these codes are

identified to help develop the themes. Themes are broader aspects, and several of the codes

would be combined to develop a single theme. Thus, the benefits of ML and AI could be

combined along with their sub-benefits into one theme as a benefit. There could be situations in

this stage where the codes could be considered too vague or irrelevant to the questions being

explored, in which case those codes could be ignored. Once this was done, we reviewed the

themes to see if they accurately represented the context mentioned in the study. Initially, for the

theme of Benefits, we had individual benefits for AI and ML, and after review, it was decided

that they could be grouped.

In some cases, we could also split the themes, but mostly, there were no issues, and the

theme review showed all the themes would be sufficient and helpful in achieving the research

objectives. We move to the next stage, where these themes are named based on the context they

best describe. Defining the themes would involve formulating what the theme means and

figuring out how this would help better understand the data while naming, as the name suggests,

creating a name that is easily understandable and understood. The themes were developed ,

reviewed, and re-developed as required based on the findings. The reviews allowed us to see if

any concepts were missed and if these codes needed to be revisited. After the completion, a peer

review was undertaken to see if the themes identified are valid and reliable and match the

interpretation and to avoid any bias. In the process, two colleagues helped, and each of them had

their own perspectives on the themes and recommended creating a separate theme on decision-

making as they felt the topic was significantly mentioned and is critical to the overall study as

this has been highlighted as a significant benefit and reason behind the adoption. The decision-

83
making aspect is one of the critical concepts when reviewed; hence, this was reviewed and added

to the initial themes list. Thus, based on the thematic analysis, we identified the critical aspects

and explored them in the coming sections of this chapter.

It is to be noted that in the responses, there are situations where the respondents have

mixed AI and ML and used them interchangeably. This is mainly due to the perception and view

seen in literature where AI and ML are considered to be the same, and to a larger extent, this is

true. Literature also has highlighted that while this is true, ML is just a small part or subsection

of AI and does not cover the entire capability of AI systems. The terms are closely related and

often interchangeable in casual conversation, media, and marketing materials. The justification

by the media for using these terms interchangeably has been that it allows to simplify the

complex technical concepts for the broader audience and also simplifies it for those with less

experience dealing with such tools. The study's introduction initially described a definition of

Machine Learning and what constitutes ML, how it differs, and examples of how it was used in

different fields, including project management. In this way, the participants were informed about

the study's purpose, and questions on artificial intelligence were included to test their

comprehension of these distinctions. However, some instances might still exist in the

communication, which is based on the overall communication and lack of technical expertise in

the field as most respondents are not tech-related but work as project managers across different

fields.

Ethical Considerations

One may argue that no scientific project is complete without first addressing ethical

concerns. The green light from the IRB and the participants' assent are two of the most crucial

requirements. Reviewing the proposal before research begins, ensuring that participation is

84
voluntary, and providing potential subjects with the information they need to make informed

decisions are all part of the IRB's mandate to protect human beings. Everyone who took part in

the study did so voluntarily, and they all gave their informed consent before the interview ever

began. All of them were informed regarding the research and its requirements and why the

research was being carried out, like the nature of the study, intent, and synopsis of the content.

We made sure to let each participant know that they could always stop the interview at any point

if they didn't want to be a part of it. They are also free to ignore questions they feel

uncomfortable with. There was no vulnerable population in the study.

The researcher will not collect the study participants' names or other identifying

information. The participants will be informed that they were assigned a random tracking

number to protect their privacy. The researcher will create an alphanumeric pseudonym, such as

P1 and P2, to replace the participant's name before the interview. The researcher will not include

any names of companies but instead identify the company using the aforementioned

alphanumeric code, which would mask any company details.

Summary

The research was a generic qualitative approach using an open-ended interview approach

with a focus on experts in the field to provide answers to the established questions. The

interviews and personal experiences would help validate the benefits and understand the

technology's current use. The chapter also established that 4-12 participants were the ideal

research method and a thematic or content analysis was identified as an optimal tool for the

analysis. The chapter also covers the consideration taken to ensure validity was provided. A set

of guidelines that were followed to ensure ethical research practice was also listed in the chapter.

85
Thus, with these, how data would be collected and analyzed are established, and the next chapter

will focus on providing the results and analysis.

86
 CHAPTER 4. RESULTS

Introduction

The chapter focuses on providing the results that were identified based on the analysis

carried out for the data collected. The chapter will also cover if there is any enhancement that

was achieved in the actual data collection from the methodology that was conceived in Chapter

3. The data collection results reflect the interviewers' question, their interpretation, integration,

and synthesis of the feedback received after each round. The research is used to study the

machine learning adoption in project management through AI practices, how these influence

project management, and the benefits to be gained from its adoption in the future.

Data Collection Results

The researcher interviewed individuals from different walks of life who specialized in

project management. These individuals worked in different industries and had different

experience ranges. That said, the experience of the participants was over ten years. The interview

code for each participant, experience, and industry is in the table below.

Table 2

Participant Demographics

Interview Code Experience Industry

IN01 15 years total Construction

87
 9 to 10 years in the

current field

IN02 12 years IT

IN03 25 years total Marketing

10 years in the current

field

IN04 18 years Software development

IN05 22 years Pharmaceuticals

IN06 21 years Automotive

IN07 19 years Manufacturing

IN08 22 years Finance

IN09 11 years Networking (IT)

IN10 21 years Construction

14 years as a project manager

88
 As we see, the average experience of the participants is 17.5 years. Only for the IT and

construction sector did we see two representations, while for all others, there was just one

representative, meaning eight industries were represented. Of the total ten, the projects and work

undertaken by each were individual though all of them worked as project managers. This wide

variety allows us to understand the variety in the requirements and activities and highlights the

level of AI use in each industry or project type. The lowest experience was 11 years by IN09,

while the highest was 25 years. This large range of experience shows the wealth of knowledge.

This knowledge would be critical in highlighting how the situations have changed.

Each individual has had successes and failures, with most respondents highlighting their

success in project management more than their failures. Each of the respondents had a different

level of involvement and project complexity. IN09, who has worked in diverse industries,

highlights the difference in complexities with their experience working on two distinct projects.

One was a project for a hotel chain that was to establish a wireless network for the hotel chain. In

contrast, the other project was a government institution focused on security evaluation. The scale

and complexity of the project significantly differ between the two. IN01 responded more in

terms of why projects would succeed or fail and highlighted the role of the different

stakeholders. His comments are as follows:

Some projects are run effectively and smoothly; it is quite surprising to see the smooth
operation. I have been a part of a few of these, and I see the difference being that of the
project leader. When the project leader is effective, the project runs extremely smoothly.
In the later years of my work, I have had experience leading projects, and I still do. I pick
up lessons from these projects that were run smoothly and what the leaders did well and
try to adopt them. I have had some issues with the projects but, overall, have been able to
complete the projects. Some key points I learned were the need for effective
communication and planning and using past information and knowledge to identify risk
areas to manage the risk better.

89
 When we explore the response of all respondents, we see that each of them has had

success as a project manager, but the definition of success could be different, as highlighted in

the literature. Regardless, this shows the wealth of experience these individuals have and

highlights the respondents' reliability in providing insights into the topic.

Data Analysis and Results

Based on thematic analysis of the interview responses, we identified the following

themes that are critical and would help in answering the research questions:

• Integrated into the current system

• Usefulness and criticality

• Limitations that influence adoption

• Decision-making impact

• Benefits

• Increasing acceptance

We explore these themes in depth in this chapter and explore the comments and differing

opinions among the individuals. Table below provides a summary of the themes and key codes

that were identified in developing these themes.

90
Table 3

Codes and Themes

Themes Codes

Integrated into the current system Used, resource allocation, scheduling,

management, project management

Usefulness and criticality Efficient, fast, data volume, insights,

flexibility

Limitations that influence adoption Compatibility, identifying right

system, cost, change resistance

Decision-making impact Improved decision making, more

informed, faster

Benefits Quicker decision making, speed,

flexibility, effectiveness, efficiency,

larger information base, better

planning

Increasing acceptance Support, guidance, cost reduction

91
 It is also important to note here that there are many concepts or codes that are found to be

common or find presence in other themes, and the below chart shows the relationship between

the different themes and components, highlighting how all these factors are influencing the

themes and highlights why some concepts are repetitive across the themes.

Figure 11

Commonalities and interlinks between the themes

Theme 1: Integrated into the Current System

The findings from the interview show that AI-based systems have been integrated

effectively into project management practices across various industries. IN05, who is from the

pharmaceutical industry, mentions that AI systems are used in project management, focusing on

three aspects: resource optimization, decision-making, and risk management.

92
 Consider a pharmaceutical venture that entails the creation of a novel medication. This
project involves a sizable team of scientists, researchers, and other stakeholders, and its
completion is anticipated to take several years. This endeavor is fraught with dangers,
including the potential for negative patient outcomes, delays in regulatory approval, and
cost overruns. The project manager uses AI tools to examine project data to mitigate
these risks proactively. The use of AI techniques allows for identifying possible dangers
and predicting their propensity to materialize. For instance, AI techniques may determine
that a change in the regulatory environment is highly likely to cause a delay in regulatory
approval. The project manager can then proactively take action to reduce this risk, for
example, by allocating more resources to quicken the regulatory approval procedure or
by altering the project's schedule. The project manager can also optimize resource
allocation by using AI. The AI technologies may examine project data to spot future
resource shortages and make recommendations for how to allocate resources best. For
instance, the AI tools might advise the project team to devote more resources and
concentrate on a certain aspect of the project that is essential for regulatory approval.

In manufacturing, the response from I07 highlights the capability of AI systems to

optimize operations.

In a different assignment, we were charged with streamlining a food processing facility's

production process. The project's objectives were detecting and decreasing production-
related waste, enhancing product quality, and boosting production effectiveness. In order
to examine production data, spot patterns, and offer suggestions for process
enhancements, we deployed AI algorithms. The AI algorithms also gave us the ability to
anticipate possible process disturbances and take preventative action to lessen them. The
client was able to cut waste by 15% and boost production efficiency by 20% as a result of
the project. Overall, by reducing costs, improving quality, and increasing productivity, I
have been able to offer my clients effective results because of my experience utilizing AI
in project management.

Overall, we see that AI systems are widely adopted with algorithms used in decision-

making, monitoring of risks, and even optimization of project activities to gain the required

efficiency. In addition to the focus on AI, we specifically asked the respondents about the role of

the technology. IN02 highlights how it has been used when it comes to the IT sector.

Future forecasting is one of the most popular uses of ML, where ML algorithms can
examine data from previous projects to estimate future schedules, costs, and resource
allocation. In order to identify prospective risks based on past data and flag them for
project managers to address, ML can also be utilized in project risk management.
Software testing is another area where ML algorithms can be used in IT project
93
 management to automate testing procedures and increase the reliability of test results.
ML can also be employed in project collaboration and communication systems to
enhance team productivity and collaboration. Overall, using ML in IT project
management can result in more effective decision-making, better project planning, and
higher success rates.

From the response, we can deduce that algorithms are employed by IT firms for the

purpose of project scope and planning optimization, risk management strategy development, and

historical data analysis for the purpose of estimating and identifying potential obstacles.

Collaboration and information exchange among stakeholders are both enhanced by the usage of

ML-based algorithms. The team can make better judgments with the large amounts of data at

their disposal. IN10, on the other hand, lists the various areas in which they have used machine

learning like in resource management, which includes predictive maintenance and resource

distribution. IN10 also highlights the capability of machine learning when it comes to identifying

and tracking the risk associated with the projects, updating the risk register, and planning better

to tackle these risks, which has been attributed to the large amount of data available for analysis.

Machine learning can analyze data from a variety of sources, including weather reports,
labor statistics, and data on the performance of equipment, to identify potential hazards
and assist project managers in taking preventative actions to avoid them. By examining
data from sensors, cameras, and other sources, machine learning can, for instance, assist
in identifying potential safety concerns on a construction site.

IN10, with experience in the construction sector, is said to highlight risk management

where factors like the weather could significantly impact the project and are critical risks to the

operation and success. Regardless of the industry, we can see from the individuals' responses that

using machine learning like AI provides the same benefits. The machine learning algorithms

allow for improvement in project management in various areas, including developing project

scope, risk management, and communication. Improved scheduling and improved decision

making is the hallmark when it comes to the use of AI and ML in project management in various
94
industries. We see that most companies have adopted and integrated these practices and have

experience. That said, some companies still have not adopted these practices and are slow. While

in some industries like IT, the adoption is quicker, in other areas, the adoption is slower and is

based on cost and need basis.

Theme 2: Usefulness and Criticality

Every response in the study has highlighted through multiple questions that significant

benefits are gained from using AI techniques like ML in project management. While AI has been

integrated into multiple aspects of different operational activities, the tool's capability in better

project management has been highlighted as a significant factor for the organization. When

asked about the usefulness of AI in project management is considered to be highly useful, and

this is based on the perspective of different individuals from different industries. IN09, IN04, and

IN02 could objectively be considered to have the same requirements in the project. Software

development, for example, is extremely complicated and often many project management

models have been used to simplify and improve project management. It has allowed flexibility

and making improvements in the software to ensure that the final product meets all the

requirements while removing all the issues with the software. The usefulness of ML in project

management can be inferred from the questions on AI and the benefits they have seen through

the use of ML in their project management. From the response, we can see that all the

respondents truly believe that, irrespective of how AI or ML is used in the company, the tools

have helped improve project management practices. Below we provide a table that highlights the

different participants and their perception of the usefulness of AI in project management and

highlights the key areas they have discussed.

95
 The table shows that some concepts are common across all studies, that is, the usefulness

in some project management activities. These are risk management, project planning, and

decision-making. The intensity or the use across the different industries is quite common as these

are the activities for which AI technology has been used largely. ML, with its algorithm, as

mentioned and highlighted in the interviews, can scan through large data, even from past

projects, making the decision-making process much simpler and based on the information

mentioned by the respondents. In complex projects like software development and construction,

the capability of analyzing and having more information at hand would always help improve the

final decision. All agree that AI's potential and usefulness in project management are high, and

there is more potential for the future in the technology.

96
Table 4
Theme 2- Responses

Respondent code Do they Which areas? Comments ML Uses

consider

AI useful?

IN01 Yes Planning or “Yes, it is very useful, and I have Predictive maintenance
only seen it in two areas, but it
scheduling, can be used as long as we can (for equipment in the
provide it with the right data in
communication, other areas.” project)

and risk Resource management

management Project Planning or

scheduling

IN02 Yes Risk management, “Overall, AI is critical, simplifies Resource allocation

various activities or tasks in
scheduling, and project management, and is more Project Planning
accurate. As mentioned in the
communication previous question, we can only Software Testing
consider so much information in
decision-making when we do it
manually. Still, AI can use and
provide an opinion based on more
information as long as the

97
 information is fed to it. Thus, it’s Communication and
a more informed decision, and it
provides suggestions and does not stakeholder
make the final decision. It
simplifies and allows us to management
understand the different scenarios
we might have overlooked if done
completely manually.”
IN03 Yes Risk management, “Absolutely. Based on my own Stakeholder
experience, I think AI is really
resource allocation beneficial for project management management
and has the ability to change how
planning, and we handle projects completely.” Predictive forecasting

communication. “For instance, as part of one of Personalization

our most recent marketing
initiatives, we handled routine
stakeholder contact by using an
AI-powered chatbot to address
frequently requested queries and
deliver project updates. This
chatbot was linked to our project
management application, enabling
it to instantly reply to stakeholder
questions and automatically pull
project-related data. Our project
managers used to spend a lot of
time answering stakeholder
questions, which was a time-
consuming and frequently
repetitive activity before we
started utilizing the chatbot. But
98
 with the chatbot in place, we were
able to save a substantial amount
of time and money because it
could deal with simple questions,
freeing up our project managers to
work on more difficult ones.”
IN04 Yes Monitoring, “I think AI has a lot to offer Predictive Maintenance
project management based on my
Resource experience as an IT project Quality controls
manager.”
Allocation, E-commerce
“AI can offer insights into project
Schedule status, spot possible problems, recommendation
and assist project managers in
Optimization making defensible choices about engines
how to enhance project results by
evaluating project data in real- NLP
time.”

IN05 Yes Quality control “I have observed how Drug discovery:

significantly AI may affect project
(risk management), management. By offering insights, Supply Chain
forecasts, and recommendations
optimization of that are not achievable using management
conventional project management
operations methodologies, artificial Personalized Medicine
intelligence [AI] has the potential
(resource to transform the way completely,

99
 allocation), we approach project management. Quality and Safety
AI can assess project data from
decision-making the past and forecast results for Monitoring
the future. I've analyzed project
data from the past using AI to
make predictions about the future.
For instance, to estimate the entire
timetable for a clinical study, we
employed an AI algorithm to
assess data on patient recruitment,
trial length, and resource usage.
As a result, we could allocate
resources more wisely and ensure
we had everything we needed to
finish the trial on schedule.”
IN06 Yes Risk management, “I can say with confidence that AI Autonomous driving
has the potential to be very
scheduling, and helpful in many aspects of project optimization and
management because I have
resource allocation experience in this area.” development

“Artificial intelligence [AI] can be Performance

used to track and examine data
from numerous suppliers and spot monitoring and quality
possible bottlenecks or supply
chain interruptions. In order to control
guarantee that materials and parts
are supplied on time and at the Maintenance planning
proper price, it can assist project
managers in making more
informed decisions about sourcing
and supply chain management.
100
 Artificial intelligence [AI] can be
used to track and analyze data
from multiple vehicle systems and
spot potential flaws or upkeep
requirements early on.”
IN07 Yes Planning, risk “I am absolutely certain that AI is Quality control
a really beneficial tool for project
management, management.” Monitoring project

project monitoring, “AI algorithms are able to monitor equipment

projects and deliver real-time
and decision- updates on important performance Resource planning
indicators [KPIs]. Large-scale
making. projects, where there are several
moving pieces and it can be
challenging to keep track of
everything, have found this to be
very helpful. AI has shown to be a
really beneficial project
management tool.”
IN08 Yes (more Forecasting and “I think artificial intelligence [AI] Security and reliability
has a lot to offer in project
needs to be optimization of management. Project managers (Risk Management)
can benefit from AI tools like
done tone) resource predictive analytics, machine Project Planning
learning, and natural language
allocation, learning processing to evaluate data more Stakeholder
quickly, spot possible problems
from the past (risk and risks, and enhance management
stakeholder involvement and
communication.”
101
 management), and

decision-making.

IN09 Yes Data analysis and “I think artificial intelligence has Risk Management
enormous potential for project
decision-making management, especially in the Project Planning and
areas of data analysis and
(resource decision-making.” organizing

allocation and “AI, for instance, can examine

past data and offer
scheduling) recommendations on the best way
to allocate resources or the best
way to schedule projects.”
IN10 Yes Risk management, “I think AI has a lot of promise to Predictive Maintenance
improve project management in
decision-making, the construction sector. The Risk Management
construction sector has generally
resource been sluggish to adopt new Resource management
technology, but as demand for
allocation, and sustainable and energy-efficient Quality controls
buildings rises, more businesses
improved site are beginning to look into the Planning
potential advantages of AI.”
security.
“AI can also contribute to
increased site security. Artificial
intelligence [AI] systems can
identify possible safety issues and
recommend countermeasures by
evaluating data on accidents and
102
near-miss incidents. On
construction sites, this may help
to prevent accidents and injuries.”

103
 When it comes to criticality for project management, the above factors are also

highlighted again and thus justify the criticality of machine learning for project management.

When we look at IN09 mentions that ML is critical for project management and its success, they

highlight the same usefulness explained before as factor that makes the tool critical. This

usefulness is said to help optimize or enhance the project performance while lowering the

expense based on their experience.

IN09 – By offering insights into project performance, spotting potential hazards and
opportunities, and predicting project outcomes, machine learning (ML) systems can assist
project managers in making better decisions………machine learning systems can aid in
resource allocation by making sure that the appropriate resources are assigned to the
appropriate tasks at the appropriate times. This can lower expenses and increase project
performance.

This view is shared by others too, and they too focus on how the ML has been useful in

their project management and highlights them being critical based on the success they have

achieved as the result of the integration of the tool. IN05 is another example where more in-

depth experience is provided.

IN05 – I believe that the importance of machine learning (ML) in project management for
pharmaceutical projects is rising. Numerous case studies already in existence show how
valuable ML is for managing pharmaceutical projects. For instance, ML can be used to
evaluate vast volumes of data, such as the outcomes of clinical trials, to find trends and
insights that might guide decision-making in the drug development process. By
identifying the compounds that have the highest chances of succeeding in clinical trials,
ML can also be used to accelerate the drug development process. Drug safety monitoring
is a further area in which ML might be useful in pharmaceutical project management. By
examining data from numerous sources, such as adverse event reports, electronic medical
records, and social media, ML can assist in identifying potential safety risks.

Overall, the interview responses agree that ML technology is useful and critical for

organizations and the industry's future.

104
Theme 3: Limitations that Influence Adoption

Before exploring the limitation, it is critical to note that to be adopted to the process,

there could be some limitations for the ML, system, or the implementation holding the successful

adoption. It is important to note that these could also be termed as barriers as they restrict or limit

the ability to use these tools in project management. Hence in this theme, we have explored the

limitations as barriers hindering its widespread adoption. For example, cost could be considered

to be a barrier. Still, the higher cost is also a limitation, as not all companies are willing to spend

the same amount of money on the product, which would hinder its acceptance. Based on the

interview responses, we have the table below that represents the different barriers to ad option.

Table 5

Limitations mentioned across interviews

Interview Codes Limitations and challenges

IN01 Cost, lack of other resources, Regulation

standards, data quality, and quantity.

IN02 Cost, data security, and regulatory controls

IN03 Lack of clear guidelines, finding the right

solution, lack of technical knowledge, and the

complexity of the systems

IN04 Awareness and trust in technology, Bias in

algorithms, data quality and quantity, security

threats and privacy standards, and integration

into the existing process

105
 IN05 Lack of specialized data, meeting standards

set by government agencies, Privacy

Concerns

IN06 Lack of qualified Employees, Cost and

specialized nature of infrastructure and other

resources, Security and safety of data,

regulatory controls.

IN07 Integration to existing systems, cost, and

uncertainty of the investment

IN08 Regulatory concerns regarding data and

standards for operation.

IN09 Lack of qualified personnel, data quality and

availability, integration challenges with the

current system, cost, and resistance to change

IN10 Lack of awareness and knowledge, Cost, and

Need for more sophisticated decision-making

abilities.

When we look at Table 2, we can see that some limitations are repetitive and common

across all interviews, but some are unique or mentioned in just a few. For example, IN10

mentions the need for more sophisticated decision-making abilities, which infers that the current

algorithms are not quite enough for the company and its activities and may be the same in some

of the companies within their industries or across specific niche industries based on the
106
complexity of their projects. While these are important findings for the research, there are also

common factors we need to address.

Cost and other resources: The overall cost of implementing and integrating ML

into their applications for project management is a major concern as not all organizations have

the required resources.

IN02- “It can be expensive for many firms to implement ML because it needs
investments in technical infrastructure and the hiring of qualified personnel”.

Considering the technology is new and requires additional infrastructure support

and qualified individuals who can operate the system makes it a significant investment that

companies think twice before using this unless they can justify the cost and are aware of the

value the tool would add. Individuals with experience in the field are also very few, as the

technology is still relatively new. People who understand the system and its application would be

critical for smooth integration and finding the right system for the company. The cost factor is

not just linked to hiring people and developing the required hardware and infrastructure. Still,

there are also decisions and costs involved in data storage and security, data itself, tools and

software development, and system maintenance and update to ensure technology improvements

and safety. This creates a sense of uncertainty among the organizations and acts as a barrier.

IN07 – The uncertainty around return on investment serves as another barrier to adoption.
Companies want to make sure that the investment pays off in the form of enhanced
efficiency, decreased costs, and higher quality because ML deployment demands
substantial time, resources, and financial investment.

The challenge of integrating the system with the existing infrastructure is critical to the

smooth running. If the system cannot be integrated with the existing systems, it could require

upgrades or replacement of the hardware and software systems, increasing the project cost. There

107
is also the issue of what ML application is ideal. This concerns whether to follow existing

software or new applications customized for the organization to ensure that the application

operates based on company demands. The cost here, too, would be different, and the thought or

consideration is only given when there is a clear understanding and knowledge regarding the tool

and its capabilities or guidelines. There is also an issue that has been addressed in the interview:

even with effective training, these systems are complex and require expert help and technical

competence to ensure they run smoothly. Employees who are not trained properly or are

unwilling to learn the system might find it difficult to operate and get the full benefit. Employee

training would help them better understand the system and allow them to be better prepared to

use it, but the system is still complex. Thus, specialized training or experienced individuals

would be required to tackle some of the complex aspects of the process. This would also

highlight the importance of having customized applications that would allow the companies to

request customized interfaces that would simplify the access and use of the system. At the same

time, experienced professionals or contracts could be given to train and maintain the systems.

Regulatory concerns and security: This is a significant external factor that could

significantly impact the organization and its image of the company to the outer world.

IN01- Regulational obstacles are another problem. To ensure compliance and safety, a
number of rules and guidelines are applicable to construction projects. Implementing ML
might be difficult in this setting because the algorithms and data must adhere to legal and
ethical standards.
As mentioned in the interview response from IN01, specific guidelines are critical for

each industry that needs to be followed, which, if the company does not follow them, would

invite fines and other issues that could significantly impact the company. It is also critical for us

to see that depending on the industries, the standards would differ, or the regulations for the use

of technology in their practices would differ. Still, the common challenge is related to data
108
security. Established protocols, standards, and privacy issues all act as barriers. Data that is

collected in some industries have significant privacy concerns like health and finance.

Precautions are taken to ensure the security of the data. However, if we were to utilize this data

for other initiatives, it may compromise customer privacy and pose a security risk. Any loss or

abuse of this data would constitute a serious breach of trust. The EU and other governments have

been mulling over regulations and control of the organizations on the use of data. This causes

companies significant concern about how the data would be used and if the company will

comply with the established standards. These worries create doubt and act as a barrier to

adoption.

Data: The data required for training the system that would help provide the required help

is vast, and without a large amount of good quality data, the ML system will not be capable of

providing the required assistance. Not all field has measures that are used, which means the data

that might be available for companies to use could be limited. Thus the capabilities of the system

could be limited and hence is not considered viable until the data issue has been sorted. There is

also the issue of bias in the system, as mentioned by IN04.

IN04 – The possibility of bias in ML algorithms is another issue. Due to the fact that ML
algorithms are trained on historical data, they may unintentionally pick up on and
reproduce biases that exist in that data. This may result in biased judgments and support
current biases or inequities. To address this issue and make sure the ML system is just
and equal, thorough data preparation and algorithm development are needed.

These issues and concerns are critical to be addressed to ensure that these systems are

suitable.

Theme 4: Decision-Making Impact

109
 All respondents throughout different questions have highlighted the role and capability of

the ML tool in decision-making. Irrespective of the industry, ML is said to help improve

decision-making in all fields of project management. It is mentioned that ML helps identify

patterns that are often not visible from manual searches but through analysis of vast amounts of

data specific to a situation and based on project management practices that allow for improving

the information available for the decision-makers. For example, when we look at pharmaceutical

projects, as mentioned by IN05, there are many issues or challenges that need to be identified to

ensure the project operates smoothly and is a success. IN05 mentions using different genomic

and proteomic data that help identify the different therapeutic targets and help customize the

medicine for the people based on their genes. The use of ML is said to help in reducing biases

and errors that might arise as a result of this bias.

IN05- ML can optimize the drug discovery process, predict which compounds are most
likely to be effective, and analyze genomic and proteomic data to identify prospective
therapeutic targets. Based on their medical histories and genomic information, ML can
also be used to identify patients who will most likely benefit from certain medications.
ML can assist in lowering the likelihood of errors and biases while also increasing the
accuracy and effectiveness of decision-making. ML systems can spot patterns and
relationships in data that humans would miss while avoiding the influence of subjective
elements like personal biases.
The above patterns within data are said to help with better risk management, resource

allocation, and scheduling, all of which have a critical role to play when it comes to the success

or failure of a project and thus are critical. In construction, they mention how ML could be used

in forecasting and keeping track of the weather, which allows it to plan the activities better and

ensure the weather risk is negated. In this process, all the information is collected and provided

to the project leader, who needs to use it and make the right decision based on their experience.

Thus, ML has helped simplify the decision-making process, and ML is said to help humans make

the right decision though individually, these systems are not capable of making their own
110
decision. The lack of professional experience and understanding of the various aspects of the

workforce and how culture and other behavior impact the workforce making decisions is not

possible, and this is not something that ML can tackle. It requires the experienced manager to

decide as they know various factors based on real experiences and analysis of the workforce.

These factors cannot be trained into the ML algorithm at present as the system does not analyze

empathy, emotional aspects, and behavioral traits. ML can only do what it is trained to do while

understanding the people in the workforce or understanding a circumstance like maintenance

requirements that might be beyond metrics is only possible with an experienced individual.

IN10 –The application of ML can assist project managers in making more informed
decisions, which is a crucial component of project management in the construction
industry. ML algorithms can find patterns and trends in past data that may not be obvious
to human decision-makers. Better choices may be made in relation to risk management,
resource allocation, and scheduling as a result of this. For instance, ML can assist project
managers in foreseeing potential dangers and proactively avoiding them. It can also assist
in determining the best way to allocate resources and adapt schedules in response to
changes in the weather or other variables that might have an impact on the timeline for
construction.
Additionally, ML can help with cost estimation, which is essential for controlling project

budgets. It's crucial to remember that ML is not a panacea and shouldn't take the place of

professional human judgment. Although ML is a tool that can support and supplement human

judgment, crucial decisions in construction project management still require human judgment.”

Theme 5: Benefits

The below table is used to highlight the various responses regarding the benefits of the

system.

Table 6
Benefits of ML in project management

Interview Codes Benefits

111
IN01 Improved productivity, better cost control,

improved resource allocation and

management, increased safety, and the ability

to automate repetitive tasks like report

generation.

IN02 Predicting or forecasting project progress and

risks, automating the repetitive process, use

large amounts of information to improve

decision-making.

IN03 Forecasting and better trend analysis of

different data provide better insights,

automation of specific tasks, capability to

provide personalized solutions, and improved

efficiency in all activities.

IN04 Improved precision in fault detection, past

data could be used to predict or identify

similar errors and code modification, reducing

future flaws and improving the knowledge

base, Improved communication and alert

notification, improvement in scheduling and

risk management, and improved decision

making.

112
IN05 Automation of repetitive processes, improved

project timelines, efficient risk management,

and more accurate information for improved

decision-making.

IN06 Predictive or forecasting, risk management,

reducing time and cost-consuming activities

through automation, improved decision

making, and allows real-time decision-

making.

IN07 Optimizing the product or service, reducing

waste, improving efficiency in operations and

all project activities, reducing costs,

forecasting tasks and resources, and managing

risk.

IN08 Forecasting trends and risks, improved

information for decision-making

IN09 Uncovering potential hazards and

opportunities, forecasting project outcomes,

automating tedious operations, improving

productivity and reducing errors, resource

optimization

113
 IN10 Improved decision-making, analysis of

historical data, and a vast volume of data from

multiple sources that improve information at

hand, improved efficiency in all activities like

scheduling, resource allocation, and work

allocation, better quality control, automation

of communication and collaboration, and

improvement in the area

From the table, we see many benefits are found to be common. One of the main

advantages of using ML in project management has been highlighted to be the automation of

tasks done manually in the past. There are different tasks to be carried out by the project

manager, some of which are repetitive. The capability of the algorithm to automate a process like

automatically sending reminders of the deadline, developing reports, and also updating the

schedules and risk register are all situations that allow the project manager to focus just on the

various tasks where his attention is required while allowing the machine to update and do these

chores with accuracy. Considering there is no decision-making in the process, the systems can

easily be programmed. They could use the vast amount of data to analyze them and provide real-

time analysis. This allows the project manager the ease of making his decisions much quicker. In

this, some of the applications might improve the visibility and monitoring aspects of the project

along with coordination. The different applications interacting with each other and ML sending

messages to all required based on preference on the risk register and updates would allow for

improved planning or strategies to mitigate risk and thus help the system be more secure.

114
 IN04 – Finally, repetitive tasks can be automated using machine learning, freeing up IT
project team members to work on more strategic tasks. Machine learning, for instance,
can be used to automate testing and quality assurance procedures, which can shorten
development times and eliminate errors.
This automation of tasks allows the project managers to focus on other tasks and is a step

toward optimizing resource allocation, risk management, and schedule management. In the

automotive factory, as described by IN06, in traditional management, part manufacturing is done

with manual control on predictive maintenance, which heavily relies on the experience of the

people allocated with the tasks. Some of this predictive maintenance would be carried out based

on expected standards that the individuals are aware of based on their experiences. At the same

time, there could be hidden patterns or factors that are often not considered, like the difference in

life based on the material used or the speed of the process. These factors could influence the

maintenance requirement, and when there are no changes made to maintenance, check which

done manually would be spread and not on a daily basis. If we replace this task with ML, it

ensures that the system would use sensor data collected in real-time, help forecast when

maintenance might be required, and alter the timeline based on the activities undertaken and the

data collected from different sensors. This would mean that the system can avoid any unforeseen

failures, which would help reduce downtime.

IN06 – ML can aid with predictive maintenance, which is essential for avoiding
unforeseen failures and reducing downtime. In order to identify patterns in sensor data
and forecast when maintenance is necessary, machine learning (ML) algorithms can
replace time- and money-consuming physical inspections. Second, by locating flaws and
investigating their causes, ML can raise the caliber of products. ML algorithms may
uncover patterns and correlations that are difficult for people to see by evaluating data
from a variety of sources, including production processes, inspections, and customer
feedback. This enables continuous development.
Downtimes could be expensive and influence the three factors that determine the project's

success. These are cost, time, and quality. Poor maintenance or delayed maintenance could mean

115
poor quality products, while breakdown could mean replacing parts that could be costly, and

also, there would be time lost in the repairs. This is another benefit that is commonly mentioned,

which is cost saving and improvement in planning and allocation. The same aspect can be used

in resource allocation as there is more clarity in the requirements, and based on the issues of the

supply chain or forecasting of risk made by ML, the company would be able to make changes to

its plans or the use of resources that would allow them to mitigate the risk while ensuring the

optimal performance. ML, due to the large amount of data that it can analyze, which would

include past projects, provide more information that can be considered while undertaking the

different planning process and developing the required strategies at each stage which in the

traditional system might be tedious as doing manual analysis of all problems might be time-

consuming. This simplifies the process and makes it quick while providing a depth of

information not seen with traditional project management. This information allows better

planning and improved optimization of the different activities.

Forecasting capability is another benefit that has been highlighted, which is spread across

different project management activities like risk management, supply chain management,

resource management, scheduling, work breakdown, and others. We have explored some of these

in the section with examples. Risk management is highly critical, as keeping track of the risks is

critical. Risk can change its importance quickly, and the risk register needs to be updated

regularly, while some factors or parameters might be ignored when done manually. The

capability to use large amounts of data and use online sources and different data sources in

addition to internal data allows the system to analyze better and update the risk register while

also being more capable of identifying a change or shift in the risk threat level and updating the

user based on these changes. These capabilities that allow for efficient management are one of

116
the major parts behind the adoption of ML and a factor contributing to the adoption among

companies.

IN06 – ML can improve supply chain management by looking for trends and patterns in
data on demand, inventory, and logistics. This may aid in cost-saving measures, quicker
deliveries, and inventory level optimization. Finally, ML can improve safety by enabling
driverless vehicles and advanced driver assistance systems (ADAS). Massive volumes of
sensor data can be processed by ML algorithms to enable real-time decision making, such
as altering a vehicle's speed and direction to avoid crashes or navigate challenging areas.
Theme 6: Increasing Acceptance

According to the user's response, we see some resistance to adopting ML across different

industries. It has been reported that some of the major factors or reluctance when it comes to

adopting ML has been the cultural resistance or the fear of change in the work culture that can be

prompted as a result of this transformation. It is mentioned that there is significant fear in the

adoption of the technology, which is one of the major reasons for reluctance to change, while

there are also those in management that are not sure about the capability of the technology and

the value it would provide to the company.

IN08 – There may be a number of reasons why management and staff are reluctant to use
ML in financial projects. Fear of losing one's work or the necessity for additional
education and training, which can be expensive for the company, are two frequent
reasons. Fear causes cultural resistance, which could hinder the success of the technology
adoption. Some people might also doubt the value of ML or not fully comprehend how it
operates. Additionally, there might be worries about the security and privacy of customer
data as well as possible regulatory repercussions of applying ML to financial decision-
making.
Some of these factors refer to the barriers that were identified in the study. In

contrast, others refer to the cultural and structural challenges within the organization and the lack

of trust in technology. Addressing these barriers and the reluctance among the organizations is

critical in improving the adoption among the users. IN09 mentions that the key points to the

adoption are education and training, collaboration, data availability, quality, monitoring of the

117
ML models, and starting small. One of the major barriers to the adoption of ML is the cost of the

system, which makes it expensive for many companies.

In contrast, the system's inability to be suitable for all purposes and companies makes it

hard to generalize the system. There is also the issue of finding the right system for the company.

Companies often make the wrong choice, which could prove ineffective, sending a negative view

of the technology to other companies. To tackle this, education and awareness creation is critical,

and one institution that could help in this would be PMI. PMI has established Project

Management and its standards, and there are guides for the different activities and training of the

people that have led to increased awareness of the best practices. With its spread and expertise,

PMI would be able to create awareness and spread knowledge on the critical aspects of ML that

would allow for better adoption.

IN06 – The Project Management Institute, or PMI, can be extremely helpful in removing
obstacles to machine learning adoption in the automobile industry. Project managers and
teams can benefit from education, training, and certification programs offered by PMI, a
professional association for project managers, to assist them in acquiring the skills and
knowledge required to make the most of machine learning technology. In order to
promote best practices and successful case studies, PMI may also help industry
professionals and experts share knowledge and collaborate. Conferences, webinars, and
other networking opportunities can help with this. In order to address concerns about data
privacy and security, PMI can also promote the creation of industry standards and
guidelines for the moral and responsible application of machine learning in automotive
projects.
Collaboration between PMI, ML developers, and different industries might be effective

in making custom solutions like customized project management guidelines provided for the

construction industry by PMI. PMI could use its contacts to better the communication and

collaboration between the two sides and find the right solutions while also documenting the

process and creating guidelines on what factors need to be considered. Once the barriers have

been addressed, the respondents believe that the technology provides significant benefits, and

118
their experience itself is proof of this and would be capable of improving the project

management process and its efficiency and could help reduce the project failures to a certain

extent. Many believe that technology is the future of the industry. As the stories of success

spread more and the barriers are removed, adoption could see an increased pace in the future,

irrespective of the industry.

Summary

The analysis of the interviews has provided specific themes that highlight the role of ML

in various industries regarding project management. Using thematic analysis, we found that the

common themes include: Integrated into the current system, Usefulness and criticality, Barriers

to adoption, Decision-making impact, Benefits, and Increasing acceptance. The themes highlight

the presence and use of ML, what the individuals perceive as the benefit and use of ML in

project management, what they perceive as the barriers that lead to reluctance in adoption among

the companies, and what steps can be taken to tackle them. The findings show significant

benefits of using ML and that it has been widely used in different industries under different

capacities like communication, planning, and risk management. Many barriers form a big part of

the reluctance, and the interview from its responses also highlights some steps that can be taken

to improve its acceptance. That said, from the interviews, it is quite clear that ML is effective in

all industries for project management. With the help of improved customization and more

specific development of applications, ML could be widely adopted and is the future of project

management.

119
 CHAPTER 5. DISCUSSION, IMPLICATIONS, RECOMMENDATIONS

Introduction

The previous chapter provided the results based on the thematic analysis carried out

during the interviews. This chapter's findings or discussion focuses on answering the primary

research questions. Also, it helps identify more in-depth findings that could help address

different business concerns and tackle these problems head -on. The study's main purpose was to

understand the perception of project managers, the perceived usefulness of Machine Learning

within the industry, and providing measures that could boost its adoption in the industry based on

the perceived usefulness. While ML algorithms have been shown to help in different project

management activities, a lack of studies focused on its impact and why such technology is still

not adopted on a wider scale in all industries. This study would help in identifying the

challenges. It would also help develop recommendations based on these findings that would help

change the business perception and help them make the decision. It would also help the

policymakers and institutions take steps by identifying the limitations expressed with the study

that would aid the companies in adopting the practices while establishing standards and

guidelines.

Evaluation of Research Questions

The research had only one research question:

What is the perceived usefulness of AI/ML across industries in project management?

The research question is focused on understanding the perceived usefulness of adopting

the machine learning or AI system in project management. Based on the data collected from the

120
interviews, talk about how machine learning is used in their organizations belonging to different

industries and highlight their perception of the benefits of using ML. These factors are said to

help highlight what influences the perception, it shed light on how widespread ML is in various

industries and on understanding the roles these tools play in their companies. It was observed

that irrespective of the industry, all the Project Managers highlighted using the ML algorithms in

different practices like risk management, scheduling or planning of projects, communication, and

other activities. This highlights that technology is being integrated into the process, and even in

an Industry like Construction, which is often considered to be a reluctant adopter of technology,

some companies have taken measures to adopt and use the new technology to improve their

project management. The flexibility and capability it provides have been highlighted in the

responses, with the respondents mentioning the multiple benefits that the companies have

achieved through the use of ML and also mentioning it as the future of project management. At

the same time, these respondents also highlight that multiple factors often act as barriers that

would hinder the adoption.

One of the primary concerns is the lack of guidelines or understanding among the

industries regarding the benefits and advantages of the ML system. While the technology is also

costlier, the companies would be willing to purchase it if they were more aware of it and had

clear guidelines for selection and its use in project management. It is to be noted here that while

some industries have the advantage of being more aware of the capabilities and benefits of ML,

not all industries would have the same advantage. In addition, while many studies highlight the

capabilities and benefits and show the successful adoption of ML, many companies have

experienced system failure due to it not fulfilling its objectives as envisioned by the firm. One of

the main attributes of the failure has been the inability to identify a suitable system. The

121
participant's responses to the study highlight the need for a customized solution for the industry's

requirements, and the technology that might be capable in one industry might not find the same

level of usefulness in other industries. It could be achieved with a common guideline or for a

generalized set of practices. That said, for industry-specific or company-specific solutions, no

single system can be adopted as is and would require the company to customize, or create a

unique solution. The cost involved in this process is a deterrence factor that creates doubt in the

minds of companies to think if the investment is worth it. While the generalized system can be

used, it is critical to make people or the company aware of its limitations and highlight where

these systems can be used. There needs to be a clear understanding regarding the scope of these

systems, what objectives or requirements would not be covered in their option, and where

customization would be required.

The study was also able to highlight some of the possible solutions for this issue: the need

for PMI and other similar institutes to work with industry experts to develop recommendations

and guidelines that would highlight the scope and requirements for each industry that would help

develop a generic system. More awareness would also need to be created, which, per

respondents' suggestions, could be achieved through regular collaborations, meetings, and

conferences. Using these methods with reputed institutions and experts to provide awareness

would help drive interest in the ML system. It could help remove any doubts or concerns the

companies have regarding the adoption. As demand for the product increases and more robust

generic and custom solutions are available, the cost could decrease, making it more affordable

for most companies.

Another critical issue regarding the adoption in some industries has been the strict

regulations and policies, along with the lack of data. While this is true for some industries, it

122
might not be true for others. While industries like Pharmaceuticals and Financial industries have

a plethora of data that they could use and are stored, they also fear how it could impact their data

security and are reluctant to use the said data. In the pharmaceutical, healthcare, and financial

industries, the data storage, the type of tools or guidelines on what type of AI or ML systems

could be used and how they are to be implemented, and in what processes they are allowed are

all governed and controlled to ensure safety and reliability. While the ML would be used for PM

activities, there is the risk of them using the data that is stored based on past projects and other

data that could be private and against the regulations or policies to use. While this is on the side

of policy and regulations, the companies themselves are worried about if the data collected and

stored would be leaked and if that could impact the trust in the companies. The companies are

more worried about the challenges it would introduce that could harm the company's reputation.

This problem also requires the experts' intervention as it would help create guidelines that

integrate the policies and guidelines to ensure the safety of the data use. Virtualization could also

be used to separate the project management data from the other functions and only access

specific operational data required for the project.

There are also areas where there is a lack of quality data that would allow for getting the

required training and thus allow for improved accuracy. Without quality data, training the ML

algorithms would be difficult. If the data quality is poor and does not cover every aspect of PM

activities, it would not be able to accurately predict risk or plan the resources, thus making the

system not beneficial. Thus, developing the data would be critical. This might require the

collaboration and digitization of existing old data that would provide more information that

could be fed to ML for fine-tuning. Modernizing the process and collaborating with experts to

create the data based on best practice guidelines could be another way to solve this problem. This

123
would also require a level of preparation and planning during the ML development stage, where

coordination and communication between the stakeholders would be important. Another issue

highlighted as a barrier would be the integration with the existing legacy systems. Different

companies use different machines from different manufacturers. While in project management,

this issue might seem of less importance, the truth is the opposite. Projects in different industries

are of different requirements, and it involves both internal and external agents. When we

consider the example of risk management itself, data is needed from multiple internal and

external sources, which would require information to be collected from multiple sources of

different types. Most modern machines would have tools to transmit data that could be used, but

there could also be others that might not have the same capabilities. Using sensors would help in

sorting this issue, but at the same time, integrating the data from different machines which might

be compatible and not compatible with the ML system and solving this could increase the project

cost and hence act as a barrier. Improved security policies for networking between the devices

could help improve the system's reliability and trust in the system. Integration generalizability

and communication protocols would be key in sorting this issue.

The response from the respondents, though, highlights many of these barriers. The main

barriers based on the solution they provided according to their perception were focused on

improving awareness and collaboration and developing guidelines. This was in line with the

findings of IPMA (2020) list and further helped expand on these views. The responses also

agreed with the findings of el Khatib and al Falasi (2021), Belharet et al. (2020) and Dam et al.

(2019) who stated that the major improvement achieved would on speed and effectiveness of

decision making through the use of data analysis. Other benefits explored in literature that

focuses on risk predictability, resource management like mentioned by Wu et al. (2014),

124
Vanhoucke et al. (2016), Munir (2019) have all been found to be true and in line with the actual

practices in the industry. Thus, from this, we can conclude that the lack of guidelines or generic

PM ML suitable for the industry and the lack of awareness of the system and its capabilities play

as the main barriers or limitations to adopting ML practices in project management.

Fulfillment of Research Purpose

The focus of the study was on understanding what the barriers were to the adoption of

ML in project management within different industries. With the findings that have been seen

based on her responses, we were able to identify certain factors like lack of awareness regarding

the ML system, cost, regulations, and data quality, all of which have hindered the adoption in the

industries. Some companies have adopted these technologies and have found them to achieve

specific benefits in project management with improvements in overall efficiency and decision-

making. The technology is found to be useful regardless of the industry where it is deployed.

This answers the perceived usefulness of the technology, which was the main focus of the study.

That said, the barriers mentioned above are hindrances that must be tackled. While this was not

part of the research question, gaining insight into barriers and recommendations helps make

organizational changes. This is in line with the literature that explored the different prospects of

the adoption of AI and ML technologies but also highlighted the barriers where multiple barriers

were identified (Bahl, 2022; Bodea et al., 2020; Eltawil et al., 2023; S. Kumar et al., 2021;

Morley et al., 2021; Rudolf, 2023; Shang et al., 2023; Ulrich et al., 2021). The study was able to

verify the barriers but also found that some of the barriers, like cost, could be tackled by

increasing the awareness among the people with regards to the benefits and if there are clear

guidelines on how to use the tool in the project management and having more confidence or trust

in the system that it would not fail.

125
 The research also highlighted how ML is being used and what activities have been

controlled and automated using the ML system, which adds to the literature and highlights the

capabilities and benefits that can be gained from the adoption, which could be used in creating

awareness among the people. Considering that the responses were from respondents who have

worked with ML in project management, their perception of the tool and its benefits would help

others to make better decisions. This also would mean that these individuals are aware of the

issues when it comes to adopting the ML system and can thus provide their own opinions on

tackling this issue. While the focus was on finding the barriers, the study was also able to

provide recommendations on how to tackle the said barriers, providing context for future

research and for industry experts and PMI to consider advancing the adoption of ML. The

recommendations are also in line with some that were seen presented by Rudolf (2023), where

clear guidance and control on the software available is provided, which makes it easier for the

companies to make the right choice and also highlights the need for having people aware of the

possibilities which is again linked to the need for increased awareness. At the same time, Rudolf

presented one solution that can or cannot be ideal for project management. The emphasis by the

interview respondents on the need for guidelines in the adoption of AI-based technologies for

project management from PMI and other governing bodies might be an ideal way to help tackle a

multitude of issues like cost, generalization which could bring the cost down for general

activities and clarity on the software or models that are ideal for adoption thus reducing the

overall risk of failure and improving the trust. PMI as a source could also help increase the data

available through its program for PM-associated learning. This would improve data availability

for training and thus improve the overall efficiency of ML technologies. Thus, the project has

126
met the research purpose and provided an enhanced view from different industries regarding the

barriers and possible solutions that would work commonly across the sectors.

We look at each theme and explore how it is linked with the literature covered in the

study. The literature shows that Building Information modeling, an AI technology, has been

integrated with project management. The technology provides information regarding the

structural and functional data within the 3D models, allowing for the virtual simulation of

construction projects (Irizarry et al., 2013; Khan et al., 2021; Koseoglu & Nurtan-Gunes, 2018).

It is critical to note here that the integration of this tool allows project managers in the

construction sector to analyze and visualize the impact of various factors on different project

phases. BIM is said to help identify potential issues in the project execution, improve resource

allocation, improve communication across different teams, and streamline tasks (al Hattab, 2021;

A. B. Mohammed, 2022). The construction sector is highlighted because it is considered one of

the sectors that is slow to adopt new technology or enact change. We also see that the literature

highlights that in IT and other sectors for project management, AI has been used in d ifferent

forms, like chatbots, for predictive modeling and decision-making support (Danysz et al., 2019).

Dam et al. (2019) mention the use of the technology for task assignments, risk assessments, and

progress monitoring.

This highlights what was seen with theme one, which is the integration into the current

system; we can see the literature, just like the interview, also highlights that these technologies

and systems have been in use and are being adopted, though not widely. For theme 1, we saw

that irrespective of the industry sector, AI and ML have been used for project management in

various capacities. For example, IT is used for developing effective scope and planning based on

a risk management plan, highlighting one aspect of the use from the interviews. In the

127
interviews, we also see that based on the responses of the participants from the pharmaceutical

industry, the tool is used for better resource optimization, decision-making, and risk

management. The review highlights that technology in project management spans different

knowledge areas and is not limited to one specific area. That said, one of the common aspects

identified from the study is using tools for risk management irrespective of the industry. This

highlights the critical and effective role that AI and ML could play in defining the project scope

and identifying and managing risks much more effectively. As highlighted in the literature,

project scope is critical in ensuring that the project's boundaries are well-defined and have

mechanisms to react to the changes. Overall, the literature and the theme help to highlight the

growing interest and use and help add to the literature on critical areas where the integration has

helped the different sectors, which add value and could help companies in similar areas to refine

their focus to use Ai or ML for these project management activities.

Theme 2 is Usefulness and Criticality, which the response in the table has strongly

supported. This theme in the interview part highlights the multiple critical aspects that project

management has helped improve project management in their organization. The respondents

highlight AI and ML's ability to simplify different tasks and enable decision-making by

providing valuable insights based on their raw capability to analyze large data sets. The focus on

the significance of artificial intelligence (AI) in enhancing decision support is consistent with

existing scholarly research that examines the advantages of AI in various domains (Prifti, 2022;

Schwarz et al., 2015). Furthermore, the participants underscore the significance of AI in the

realm of communication and the management of stakeholders. The authors discuss the utilization

of chatbots powered by AI that are capable of effectively managing stakeholder inquiries and

promptly providing project updates. Using AI in communication and stakeholder engagement

128
aligns with existing scholarly discourse. The literature extensively explores the potential of AI to

enhance communication and facilitate cooperation within the context of project management

(Prasad & Vijaya Saradhi, 2019).

The participants in the study also acknowledge the advantages of AI in the aspects of

monitoring project development and ensuring quality control. The authors emphasize that

artificial intelligence algorithms can offer real-time updates on crucial performance metrics,

rendering them particularly advantageous in projects of significant scale. The emphasis on the

use of AI in monitoring and quality control is consistent with the existing body of scholarly

research, which extensively examines the utilization of AI for these specific objectives (IN07).

This can be seen in the literature on the use of BIM, where it has been highlighted for this use in

the construction sector. Additionally, the participants emphasize the role of AI in the field of risk

management. The authors emphasize the capacity of AI to assess historical data and provide

recommendations for allocating resources and scheduling. All these factors and uses have been

covered in all literature regarding using AI and ML in project management. However, the study

shows that regardless of the industry, some similarities and tasks can be executed based on

modifications for company requirements, which could help in future developments.

The third theme is a significant contribution of the research as it helps highlight the

limitations and challenges in adoption and verify some of these factors mentioned in literature

across different industries. Most of the concerns mentioned are said to align well with the

literature. Many respondents indicated that cost is a major concern (IN01, IN02, IN06, IN08,

IN09, and IN10). Putting AI and ML systems into practice can be expensive, particularly when

hiring the right people and purchasing the equipment and resources. This cost-related issue is

consistent with findings from the literature (Bodea et al., 2020; Rudolf, 2023), highlighting the

129
financial effects of implementing AI and ML technologies in project management. Data quantity

and quality are identified among the responders as a common limitation (IN01, IN04, IN05, and

IN09). The research (Akinosho et al., 2020; H20.ai, 2020) identifies data-related obstacles as a

barrier to AI adoption in project management, which is consistent with the difficulties associated

with data availability and quality. Several responders have emphasized regulatory controls and

standards (IN01, IN02, IN06, and IN08). They have worries regarding adherence to legal

requirements and privacy norms. These worries align with the literature, which emphasizes that

one of the challenges in implementing AI and ML across businesses is complying with legal and

regulatory standards (S. Kumar et al., 2021; Rudolf, 2023).

Respondents (IN03, IN05) note a lack of technical understanding and specialized data.

These restrictions align with the conclusions of the literature, which also address the difficulties

in obtaining specialized data and the requirement for skilled workers to adopt AI and ML (Bodea

et al., 2020; Rudolf, 2023). Several responders saw integration into current systems as a

challenge (IN04, IN07, and IN09). The literature's emphasis on the difficulties of incorporating

AI and ML technologies into current project management procedures aligns with this worry

(Güngör, 2019). Security and privacy issues are emphasized (IN04, IN06). The literature's

emphasis on the value of security and dependability in risk management is congruent with the

necessity of addressing data security and privacy regulations (Rudolf, 2023). Limitations include

the requirement for more complex decision-making skills and resistance to change (IN09, IN10).

These worries align with research findings in the literature that address the difficulties associated

with organizational transformation and the requirement for AI adoption upskilling (H20.ai,

2020). Overall, the study highlights the different concerns and validates the literature by

130
validating the views across different industries, which would help in focusing on the specific

issues to tackle one by one.

The next theme is the Decision-making impact, which also mostly aligns with the

literature findings. IN05 explains how ML is used in the pharmaceutical industry to customize

medicine by utilizing genomic and proteomic data to optimize the drug discovery process. In

addition to helping to discover treatment targets, ML also helps to minimize biases and errors

that may arise from human subjectivity. The precision and efficacy of decision-making are

improved by ML systems' capacity to identify patterns and relationships in data. This is

consistent with findings in the literature on the application of AI in pharmaceuticals for

personalized treatment and drug discovery (S. Lee et al., 2021; Maceachern & Forkert, 2021;

Quazi, 2022). IN10 provides an example of how ML helps the construction industry make well-

informed decisions. Machine learning algorithms find patterns and trends in past data that human

decision-makers might miss. This allows project managers to make wiser decisions about

scheduling, resource allocation, and risk management. Additionally, ML helps with anticipatory

risk assessment and weather-related variable response. These findings are corroborated by

research highlighting the use of AI and ML in scheduling, resource allocation, and risk

management in building projects (Pan & Zhang, 2021; Zhao et al., 2016).

These instances highlight how ML influences decision-making and how it enhances

human judgment by offering data-driven insights. This synergy between AI/ML and human

expertise in making important decisions in various project management contexts is

acknowledged in the literature. It is critical to understand that while ML can improve data-driven

decision-making, project managers' experienced, nuanced judgment cannot be replaced by ML.

Human comprehension of elements like emotional considerations, cultural dynamics, and

131
workplace behavior is still crucial for making decisions. This viewpoint is supported by the

literature, which emphasizes that while AI and ML are useful tools to assist human decision-

makers, they do not replace their expert judgment (Armstrong et al., 2014; Bodea et al., 2020).

The findings, though, align with what is mentioned in the literature and validate the role and

need for balance between AI/ML and humans in the decision-making process.

When it comes to the benefits, the study's findings are similar to what is seen in the

literature and provide more practical knowledge on the benefits of the technology within the

organization. Interviewees frequently mentioned increasing productivity through work

automation (Smith & Fressoli, 2021). Better decision-making aligns with AI's data-driven

decision-support role as highlighted in literature (Andronie et al., 2021; McGovern et al., 2017).

This is highlighted by IN02, IN03, IN04, IN05, and IN10. IN02, IN03, IN04, IN05, IN06, IN07,

and IN09 all support the usefulness of AI/ML in risk management which is also supported in

literature (Attaran & Deb, 2018). Literature also imply that automation's involvement in

lowering manual effort and enhancing efficiency—a major benefit highlighted by IN01, IN02,

IN05, IN06, and IN10—is consistent with the literature (Deranty & Corbin, 2022; Rožman et al.,

2023). Efficiency and resource optimization are highlighted by IN03, IN06, IN07, IN09, and

IN10, which finds parallel in literature (Sravanthi et al., 2023; Waltersmann et al., 2021).

The topics of better communication discussed by IN04 and IN10 align with the findings

of different impact on communication within social structure (Hohenstein et al., 2023). The

literature's consideration of AI/ML's function in tracking project deliverable quality aligns with

IN10's focus on quality control (Pillai et al., 2019; Sundaram & Zeid, 2023). The reference to

real-time decision-making in IN06 is consistent with the literature on the real-time insights and

alarms that AI/ML may provide (McGovern et al., 2017). There are no new benefits identified

132
that are not already presented in the literature, which highlights a strong correspondence between

the practical experience shared by the interviewees and the literature on AI and ML in project

management.

Theme 6, which is increasing acceptance, is the section or theme that adds the most value

to the study as it highlights and provides opinions or possible solutions to tackle limited adoption

and increase the acceptance of the tool across industries. According to user comments, there is a

noticeable resistance across industries to adopting machine learning (ML), frequently related to

cultural resistance and concern about altering work cultures. The fear of change and cultural

resistance that can impede technology adoption can be attributed to concerns expressed by

management and staff about possible job losses, the necessity of supplementary education and

training, and related expenses. Furthermore, there is a lack of confidence in the technology's

potential and worth among certain members of management. These feelings are consistent with

the study's identified barriers. They also draw attention to organizational structural and cultural

issues and a lack of faith in technology. To ensure successful adoption, IN09 highlights the

importance of education, training, teamwork, data quality, monitoring ML models, and starting

small. Finding the ideal system that meets their objectives might be difficult, though, as the cost

of ML systems continues to be a major hurdle for many businesses. Bad decisions might lead to

unfavorable opinions of technology. Respondents advise organizations to increase knowledge

and education to tackle these issues, with the Project Management Institute (PMI) being a key

player. PMI has a history of advancing best practices in project management and increasing

awareness of them. Working together with PMI can lessen obstacles. It can also encourage

interaction and cooperation between ML developers and various businesses, resulting in

guidelines and solutions specifically tailored to each. These solutions are based on the perception

133
of experienced project managers and those with experience in the field. These steps could be

considered stepping stones or initial steps in tackling the problem head -on. The findings expand

on existing literature while also validating the existing literature.

Contribution to Business Technical Problem

The research adds to the existing literature on ML adoption for Project Management in

different industries. The study highlights the different ways the technology is used and how

successful these are for companies that have adopted it. Most of the respondents praised the

technology and its use for project management. While they also recognize that not all the

systems would be as effective, they are also aware that the technology is not widely accepted at

present but do believe the technology is the future. The response from people who have worked

with the technology and highlights the benefit of the system shows that the technology is indeed

helpful and is critical for it to be adopted in PM practices to improve the efficiency and

effectiveness of the process. Considering that most projects are still known to fail to improve

efficiency is critical for all businesses. When people who have experienced the benefits highlight

the benefits, it becomes increasingly necessary to explore further and integrate the practices to

achieve the same level of efficiency or improvement in PM activities. The study also highlighted

which areas the ML system is highly effective, like resource planning, risk management, and

communication, which would allow companies to explore these options based on their existing

weaknesses. The paper increases awareness of ML systems in Project Management and helps

organizations to start their research or exploration into the use of the technology.

The study also helps highlight the existing limitations and lack of guidelines. The

perception among the respondents of the need for more guidelines and collaboration could help

create a discussion concerning the findings. Hopefully, it could get the organizations to work on

134
finding a common solution that would help in developing a unique generic model for each

industry, which could solve the problem of finding the right system and also help reduce the cost.

While dedicated solutions would require more payments, this would allow smaller companies to

adopt these practices and improve their efficiency through improved decision-making. Thus, the

paper helps spread awareness and highlights the role that organizations like PMI needs to play in

advancing the adoption.

Recommendations for Further Research

While the existing research is on the perception of individuals that have been using ML

in their project management activities, it would be interesting to get the views of individuals who

might have tried it. At present, exploring the views of the individuals who work in organizations

that have not adopted these practices and comparing their perceptions would allow us to find a

common issue and confirm if the individuals' perceptions are the same on either side. This would

also highlight the consistency of the issue. Another study might explore developing a generic

ML system for one sector. While this study highlighted some of the activities used in different

industries, the study could develop a generic ML system based on the requirements after carrying

out a separate interview specific to one industry. Using this model to test their effectiveness in

the same sector would help better understand whether such generic systems are capable and can

be used or if custom solutions are required for all industries. The study would be time-consuming

and would need experiments that need to be done over an extended period of time and thus might

need to be split and done in two studies where one would develop the model and the other would

test the model. This would further the research in the field and allow for understanding if a

generic system is ideal or what options are further needed to be considered in developing a

135
generic model. Developing such a common model that could be used per industry requirements

would help reduce the cost involved in the adoption.

Conclusions

Machine learning and artificial intelligence systems have been widely adopted across

various industries for different activities. There is an increasing amount of literature available

regarding the use of ML systems in project management, how they could be implemented, and

the possible accuracy of the system. Based on this literature gap on the use of ML and if they are

effective, and what contributes to their usefulness, we undertook research where ten individuals

working as project managers across various industries were interviewed. These individuals were

asked a set of questions that covered their perception of the technology, its benefits, the barriers

and hindrance to its adoption, and how to tackle the issues. Using a thematic analysis, we

identified that the major barriers to adopting ML practices in project management are the cost of

the technology, lack of awareness, and concerns regarding data security, quality, and regulations

controlling data use. The study was also based on these individuals' experiences and highlighted

the benefits of using the systems in various industries. The respondent as a whole does not doubt

ML's capability and believes the technology is a critical tool for the future that would help the

companies make better decisions and ensure the overall efficiency of the activities. The study,

also based on the respondents' perception, highlighted measures that could help tackle the said

issues in the adoption. The main suggestion has been the collaboration between industry experts

and ML experts, along with institutes like PMI, to develop generic models and create guidelines

related to their adoption and use. Overall, the paper helps enhance the literature on ML in Project

Management and highlights how businesses could tackle existing problems related to adoption.

136
 REFERENCES

Agarwal, N., & Rathod, U. (2006). Defining “success” for software projects: An exploratory
revelation. International Journal of Project Management, 24(4).
https://doi.org/10.1016/j.ijproman.2005.11.009

Akinosho, T. D., Oyedele, L. O., Bilal, M., Ajayi, A. O., Delgado, M. D., Akinade, O. O., & Ahmed,
A. A. (2020). Deep learning in the construction industry: A review of present status and future
innovations. Journal of Building Engineering, 32. https://doi.org/10.1016/j.jobe.2020.101827

al Hattab, M. (2021). The dynamic evolution of synergies between BIM and sustainability: A text
mining and network theory approach. Journal of Building Engineering, 37, 102159.
https://doi.org/10.1016/j.jobe.2021.102159

Alami, A. (2016). Why Do Information Technology Projects Fail? Procedia Computer Science, 100.
https://doi.org/10.1016/j.procs.2016.09.124

Al-Ghamdi, A. (2009). Investigating factors affecting customers of using Internet banking : A

comparison study between Saudi Arabia and the UK. Business.

Al-Momani, A. M., Mahmoud, M. A., & Ahmad, S. M. (2016). Modeling the adoption of internet of
things services: A conceptual framework. International Journal of Applied Research, 2(5).

AltexSoft. (2016). Agile Project Management: Best Practices and Methodologies.

Amadu, L., Muhammad, S. S., Mohammed, A. S., Owusu, G., & Lukman, S. (2018). Using
technology acceptance model to measure the use of social media for collaborative learning in
Ghana. Journal of Technology and Science Education, 8(4). https://doi.org/10.3926/jotse.383

Ambler, S. W. (2018). 2018 IT Project Success Rates Survey Results. In Disciplined Agile.

AmbySoft. (2018). IT Project Success Rates: 2018 Open Research. Ambysoft.

https://ambysoft.com/surveys/success2018.html

Andronie, M., Lăzăroiu, G., Iatagan, M., Uță, C., Ștefănescu, R., & Cocoșatu, M. (2021). Artificial
intelligence-based decision-making algorithms, internet of things sensing networks, and deep
learning-assisted smart process management in cyber-physical production systems. In
Electronics (Switzerland) (Vol. 10, Issue 20). https://doi.org/10.3390/electronics10202497

Arcidiacono, G. (2017). Comparative research about high failure rate of IT projects and opportunities
to improve. PM World Journal, 6(2).

137
Arefazar, Y., Nazari, A., Hafezi, M. R., & Maghool, S. A. H. (2019). Prioritizing agile project
management strategies as a change management tool in construction projects. International
Journal of Construction Management. https://doi.org/10.1080/15623599.2019.1644757

Armenia, S., Dangelico, R. M., Nonino, F., & Pompei, A. (2019). Sustainable project management: A
conceptualization-oriented review and a framework proposal for future studies. In Sustainability
(Switzerland) (Vol. 11, Issue 9). https://doi.org/10.3390/su11092664

Armstrong, S., Sotala, K., & Héigeartaigh, S. S. Ó. (2014). The errors, insights and lessons of famous
AI predictions-and what they mean for the future. Journal of Experimental and Theoretical
Artificial Intelligence, 26(3). https://doi.org/10.1080/0952813X.2014.895105

Astalin, P. K. (2013). Qualitative Research Designs: a Conceptual Framework. International Journal

of Social Science & Interdisciplinary Research, 2(1).

Attaran, M., & Deb, P. (2018). Machine Learning: The New “Big Thing” for Competitive Advantage.
International Journal of Knowledge Engineering and Data Mining, 5(1).
https://doi.org/10.1504/ijkedm.2018.10015621

Bahl, M. (2022). Artificial Intelligence in Clinical Practice: Implementation Considerations and

Barriers. In Journal of Breast Imaging (Vol. 4, Issue 6). https://doi.org/10.1093/jbi/wbac065

Baker, S. E., & Edwards, R. (2012). How many qualitative interviews is enough ? National Centre for
Research Methods Review Paper. https://doi.org/10.1177/1525822X05279903

Barocas, S., Hardt, M., & Narayanan, A. (2020). Fairness in Machine Learning Limitations and
Opportunities. In Book.

Barreiro-Ares, A., Morales-Santiago, A., Sendra-Portero, F., & Souto-Bayarri, M. (2023). Impact of
the Rise of Artificial Intelligence in Radiology: What Do Students Think? International Journal
of Environmental Research and Public Health, 20(2). https://doi.org/10.3390/ijerph20021589

Beck, K., Beedle, M., Van Bennekum, A., Cockburn, A., Cunningham, W., Fowler, M., & Thomas,
D. (2001). Manifesto for agile software development.

Belharet, A., Bharathan, U., Dzingina, B., Madhavan, N., Mathur, C., & Toti, Y.-D. (2020). Report on
the Impact of Artificial Intelligence on Project Management [Masters Thesis]. ESIEE Paris.

Bodea, C.-N., Mitea, C., & Stanciu, O. (2020). Artificial Intelligence Adoption in Project
Management: Main Drivers, Barriers and Estimated Impact. In Proceedings of the International
Conference on Economics and Social Sciences. https://doi.org/10.2478/9788395815072-075

Bodicha, H. H. (2015). How to Measure the Effect of Project Risk Management Process on the
Success of Construction Projects: A Critical Literature Review. International Journal Of
Business & Management (ISSN, 3(12).

138
Brown, L. (2024, January 11). The Project Management Life Cycle Explained. Invensis.
https://www.invensislearning.com/blog/5-phases-project-management-lifecycle

Brynjolfsson, E., Hitt, L., & Kim, H. (2011). Strength in numbers: How does data-driven decision-
making affect firm performance? International Conference on Information Systems 2011, ICIS
2011, 1. https://doi.org/10.2139/ssrn.1819486

Cao, L. (2017). Data science: A comprehensive overview. In ACM Computing Surveys (Vol. 50, Issue
3). https://doi.org/10.1145/3076253

Chawla, V. K., Chanda, A. K., Angra, S., & Chawla, G. R. (2018). The sustainable project
management: A review and future possibilities. Journal of Project Management.
https://doi.org/10.5267/j.jpm.2018.2.001

Chi, M. T. H. (2012). Two Approaches to the Study of Experts’ Characteristics. In The Cambridge
Handbook of Expertise and Expert Performance. https://doi.org/10.1017/cbo9780511816796.002

Chismar, W. G., & Wiley-Patton, S. (2003). Does the Extended Technology Acceptance Model Apply
to Physicians Department of Information Technology Management. 36th Hawaii International
Conference on System Sciences, 00(C).

Chou, J. S., Lin, C. W., Pham, A. D., & Shao, J. Y. (2015). Optimized artificial intelligence models
for predicting project award price. Automation in Construction, 54.
https://doi.org/10.1016/j.autcon.2015.02.006

Çivril, H., & Özkul, A. E. (2021). Investigation of the Factors Affecting Open and Distance Education
Learners’ Intentions to Use a Virtual Laboratoryi. International Review of Research in Open and
Distance Learning, 22(2). https://doi.org/10.19173/irrodl.v22i2.5076

Codari, M., Melazzini, L., Morozov, S. P., van Kuijk, C. C., Sconfienza, L. M., & Sardanelli, F.
(2019). Impact of artificial intelligence on radiology: a EuroAIM survey among members of the
European Society of Radiology. Insights into Imaging, 10(1). https://doi.org/10.1186/s13244-
019-0798-3

Conforto, E. C., Amaral, D. C., da Silva, S. L., di Felippo, A., & Kamikawachi, D. S. L. (2016). The
agility construct on project management theory. International Journal of Project Management,
34(4). https://doi.org/10.1016/j.ijproman.2016.01.007

Consultancy UK. (2020, May 27). Most Construction and Engineering Projects are unsuccessful.
Consultancy.Uk.

Cooke-Davies, T. (2002). The “real” success factors on projects. International Journal of Project
Management, 20(3). https://doi.org/10.1016/S0263-7863(01)00067-9

139
Costantino, F., di Gravio, G., & Nonino, F. (2015). Project selection in project portfolio management:
An artificial neural network model based on critical success factors. International Journal of
Project Management, 33(8). https://doi.org/10.1016/j.ijproman.2015.07.003

Cubric, M. (2020). Drivers, barriers and social considerations for AI adoption in business and
management: A tertiary study. Technology in Society, 62.
https://doi.org/10.1016/j.techsoc.2020.101257

Dahie, A. M., Osman, A. A., & Omar, A. A. (2017). The Role of Project Management in Achieving
Project Success: Empirical Study from Local NGOs in Mogadishu-Somalia. International
Journal of Engineering Science and Computing, 7(9).

Dam, H. K., Tran, T., Grundy, J., Ghose, A., & Kamei, Y. (2019). Towards effective AI -powered
agile project management. Proceedings - 2019 IEEE/ACM 41st International Conference on
Software Engineering: New Ideas and Emerging Results, ICSE-NIER 2019.
https://doi.org/10.1109/ICSE-NIER.2019.00019

Danysz, K., Cicirello, S., Mingle, E., Assuncao, B., Tetarenko, N., Mockute, R., Abatemarco, D.,
Widdowson, M., & Desai, S. (2019). Artificial Intelligence and the Future of the Drug Safety
Professional. Drug Safety, 42(4). https://doi.org/10.1007/s40264-018-0746-z

Demir, K. A. (2011). Project Success as an Evolving Concept. International Conference on Software

Engineering Research & Practice.

Deng, Y., Gan, V. J. L., Das, M., Cheng, J. C. P., & Anumba, C. (2019). Integrating 4D BIM and GIS
for Construction Supply Chain Management. Journal of Construction Engineering and
Management, 145(4). https://doi.org/10.1061/(asce)co.1943-7862.0001633

Deranty, J. P., & Corbin, T. (2022). Artificial intelligence and work: a critical review of recent
research from the social sciences. In AI and Society. https://doi.org/10.1007/s00146-022-01496-x

Dhillon, H. (2018). Iron Triangle — Triple Constraints of Project Management. Medium.

Dickson, B. U., Oby, B. O., Samuel, N. N., & Udoka, S. O. (2021). Integrating Trust into Technology
Acceptance Model (TAM), the Conceptual Framework for E-Payment Platform Acceptance.
British Journal of Management and Marketing Studies, 4(4). https://doi.org/10.52589/bjmms-
tb3xtkpi

Drury, M., Conboy, K., & Power, K. (2012). Obstacles to decision making in Agile software
development teams. Journal of Systems and Software, 85(6).
https://doi.org/10.1016/j.jss.2012.01.058

Drury-Grogan, M. L. (2014). Performance on agile teams: Relating iteration objectives and critical
decisions to project management success factors. Information and Software Technology, 56(5).
https://doi.org/10.1016/j.infsof.2013.11.003

140
El Khatib, M., & Al Falasi, A. (2021). Effects of Artificial Intelligence on Decision Making in Project
Management. American Journal of Industrial and Business Management, 11(03).
https://doi.org/10.4236/ajibm.2021.113016

Elboq, R., Fri, M., Hlyal, M., & Alami, J. (2021). Modeling Lean and Six Sigma Integration using
Deep Learning: Applied to a Clothing Company. Autex Research Journal.
https://doi.org/10.2478/aut-2021-0043

Elshafey, A., Saar, C. C., Aminudin, E. B., Gheisari, M., & Usmani, A. (2020). Technology
acceptance model for augmented reality and building information modeling integration in the
construction industry. Journal of Information Technology in Construction, 25.
https://doi.org/10.36680/j.itcon.2020.010

Eltawil, F. A., Atalla, M., Boulos, E., Amirabadi, A., & Tyrrell, P. N. (2023). Analyzing Barriers and
Enablers for the Acceptance of Artificial Intelligence Innovations into Radiology Practice: A
Scoping Review. Tomography, 9(4), 1443–1455. https://doi.org/10.3390/tomography9040115

Fagarasan, C., Popa, O., Pisla, A., & Cristea, C. (2021). Agile, waterfall and iterative approach in
information technology projects. IOP Conference Series: Materials Science and Engineering,
1169(1). https://doi.org/10.1088/1757-899x/1169/1/012025

Fayad, R., & Paper, D. (2015). The Technology Acceptance Model E-Commerce Extension: A
Conceptual Framework. Procedia Economics and Finance, 26. https://doi.org/10.1016/s2212-
5671(15)00922-3

Flick, U. (2015). Introducing Research Methodology: A Beginner’s Guide to Doing a Research

Project. In SAGE Publications Ltd (Issue 2).

Gulla, J. (2011). Seven Reasons Why Information Technology Projects Fail. IBM Corporation
August.

Güngör, D. Ö. (2019). Industry 4.0 Technologies Used in Project Management.

https://doi.org/10.4018/978-1-5225-7865-9.ch003

H20.ai. (2020). Overcoming obstacles to machine learning adoption.

Han, J., Kamber, M., & Pei, J. (2012). Data Mining: Concepts and Techniques. In Data Mining:
Concepts and Techniques. https://doi.org/10.1016/C2009-0-61819-5

Hashem, I. A. T., Yaqoob, I., Anuar, N. B., Mokhtar, S., Gani, A., & Ullah Khan, S. (2015). The rise
of “big data” on cloud computing: Review and open research issues. In Information Systems
(Vol. 47). https://doi.org/10.1016/j.is.2014.07.006

Heagney, J. (2012). Fundamentals of project management, fourth edition. In Project Management

Journal (Vol. 43, Issue 4). https://doi.org/10.1002/pmj.21285

141
Henrie, M., & Sousa-Poza, A. (2005). Project Management: A Cultural Literary Review. Project
Management Journal, 36(2). https://doi.org/10.1177/875697280503600202

Hohenstein, J., Kizilcec, R. F., DiFranzo, D., Aghajari, Z., Mieczkowski, H., Levy, K., Naaman, M.,
Hancock, J., & Jung, M. F. (2023). Artificial intelligence in communication impacts language
and social relationships. Scientific Reports, 13(1). https://doi.org/10.1038/s41598-023-30938-9

Hu, Z., Ding, S., Li, S., Chen, L., & Yang, S. (2019). Adoption intention of fintech services for bank
users: An empirical examination with an extended technology acceptance model. Symmetry,
11(3). https://doi.org/10.3390/sym11030340

Hughes, G. R., Hall, W., Pesenti, J., & Ward, P. (2017). Machine learning : the power and promise of
computers that learn by example. In Royal Society (Vol. 66, Issue January).

IPMA. (2020). Artificial Intelligence impact in Project Management.

Irizarry, J., Karan, E. P., & Jalaei, F. (2013). Integrating BIM and GIS to improve the visual
monitoring of construction supply chain management. Automation in Construction, 31.
https://doi.org/10.1016/j.autcon.2012.12.005

Jaafari, A. (2001). Management of risks, uncertainties and opportunities on projects: Time for a
fundamental shift. International Journal of Project Management, 19, 89–101.
https://doi.org/10.1016/S0263-7863(99)00047-2

Jetu, F. T., & Riedl, R. (2012). Determinants of Information Systems and Information Technology
Project Team Success: A Literature Review and a Conceptual Model. Communications of the
Association for Information Systems, 30. https://doi.org/10.17705/1cais.03027

Joffe, H. (2011). Thematic analysis. In D. Harper & A. Thompson (Eds.), Qualitative methods in
mental health and psychotherapy: a guide for students and practitioners (pp. 209–224). John
Wiley & Sons.

Johnson, S. B. (2013). The secret of Apollo: systems management in American and European space
programs. In Choice Reviews Online (Vol. 40, Issue 05). https://doi.org/10.5860/choice.40-2887

Jordan, M. I., & Mitchell, T. M. (2015). Machine learning: Trends, perspectives, and prospects. In
Science (Vol. 349, Issue 6245, pp. 255–260). https://doi.org/10.1126/science.aaa8415

Kadir, S. A., Lokman, A. M., & Tsuchiya, T. (2021). Emotional Responses Towards Unity YouTube
Videos: Experts vs. Viewers Perspectives. International Journal of Affective Engineering, 20(4).
https://doi.org/10.5057/ijae.ijae-d-20-00033

Kaelbling, L. P., Littman, M. L., & Moore, A. W. (1996). Reinforcement learning: A survey. Journal
of Artificial Intelligence Research. https://doi.org/10.1613/jair.301

142
Karim, N. L. Mohd., Hadi, A. A., Hamzah, S. A. S., Rashid, P. D. A., & Salmi, N. S. (2022). A Study
on Kitchen Equipment Adoption Using Technology Acceptance Model (TAM) in School
Canteen: A Conceptual Framework. International Journal of Advanced Research in Education
and Society, 4(1). https://doi.org/10.55057/ijares.2022.4.1.1

Katz, G., Shabtai, A., & Rokach, L. (2014). Adapted features and instance selection for improving co-
training. Lecture Notes in Computer Science (Including Subseries Lecture Notes in Artificial
Intelligence and Lecture Notes in Bioinformatics), 8401, 81–100. https://doi.org/10.1007/978-3-
662-43968-5_5

Kelepouris, P. (2023). Implementation of Artificial Intelligence in Project Management and effect in

working personnel. KTH Industrial Engineering and Management.

Kerzner, H. (2014). Project Management Best Practices: Achieving Global Excellence. In Project
Management Best Practices: Achieving Global Excellence (Vol. 9781118657010).
https://doi.org/10.1002/9781118835531

Khan, M. S., Park, J., & Seo, J. (2021). Geotechnical property modeling and construction safety
zoning based on gis and bim integration. Applied Sciences (Switzerland), 11(9).
https://doi.org/10.3390/app11094004

Kiger, M. E., & Varpio, L. (2020). Thematic analysis of qualitative data: AMEE Guide No. 131.
Medical Teacher, 42(8). https://doi.org/10.1080/0142159X.2020.1755030

King, J. (2003). Survey shows common IT woes persist. Computerworld, 37(25).

Kopper, P. (2019). Limitations of Interpretable Machine Learning Methods. Limitations of

Interpretable Machine Learning Methods.

Koseoglu, O., & Nurtan-Gunes, E. T. (2018). Mobile BIM implementation and lean interaction on
construction site: A case study of a complex airport project. Engineering, Construction and
Architectural Management, 25(10). https://doi.org/10.1108/ECAM-08-2017-0188

Kothari, C. (2004). Research methodology: methods and techniques. In New Age International.
https://doi.org/http://196.29.172.66:8080/jspui/bitstream/123456789/2574/1/Research%20Metho
dology.pdf

Kumar, A., Dhingra, S., Batra, V., & Purohit, H. (2020). A Framework of Mobile Banking Adoption
in India. Journal of Open Innovation: Technology, Market, and Complexity, 6(2).
https://doi.org/10.3390/JOITMC6020040

Kumar, R. (2017). Research Methodology. A step by step guide. Sage Publications.

143
Kumar, S., Raut, R. D., Queiroz, M. M., & Narkhede, B. E. (2021). Mapping the barriers of AI
implementations in the public distribution system: The Indian experience. Technology in Society,
67. https://doi.org/10.1016/j.techsoc.2021.101737

Kuo, W. Y., Kuo, C. H., Sun, S. W., Chang, P. C., Chen, Y. T., & Cheng, W. H. (2016). Machine
learning-based behavior recognition system for a basketball player using multiple Kinect
cameras. 2016 IEEE International Conference on Multimedia and Expo Workshop, ICMEW
2016. https://doi.org/10.1109/ICMEW.2016.7574661

Lalsing, V. (2012). People Factors in Agile Software Development and Project Management.
International Journal of Software Engineering & Applications, 3(1).
https://doi.org/10.5121/ijsea.2012.3109

Lee, H. H., & Chang, E. (2011). Consumer Attitudes Toward Online Mass Customization: An
Application of Extended Technology Acceptance Model. Journal of Computer-Mediated
Communication, 16(2). https://doi.org/10.1111/j.1083-6101.2010.01530.x

Lee, I. (2017). Big data: Dimensions, evolution, impacts, and challenges. Business Horizons, 60(3),
293–303. https://doi.org/10.1016/j.bushor.2017.01.004

Lee, I., & Shin, Y. J. (2020). Machine learning for enterprises: Applications, algorithm selection, and
challenges. Business Horizons, 63(2). https://doi.org/10.1016/j.bushor.2019.10.005

Lee, S., Lam, S. H., Hernandes Rocha, T. A., Fleischman, R. J., Staton, C. A., Taylor, R., &
Limkakeng, A. T. (2021). Machine Learning and Precision Medicine in Emergency Medicine:
The Basics. Cureus. https://doi.org/10.7759/cureus.17636

Leech, N., & Onwuegbuzie, A. (2005). The role of sampling in qualitative research. Academic
Exchange Quarterly, 9(3).

Leybourne, S. A. (2007). The Changing Bias of Project Management Research: A Consideration of

the Literatures and an Application of Extant Theory. Project Management Journal, 38(1).
https://doi.org/10.1177/875697280703800107

Lim, S. S., Phan, T. D., Law, M., Goh, G. S., Moriarty, H. K., Lukies, M. W., Joseph, T., & Clements,
W. (2022). Non-radiologist perception of the use of artificial intelligence (AI) in diagnostic
medical imaging reports. Journal of Medical Imaging and Radiation Oncology, 66(8).
https://doi.org/10.1111/1754-9485.13388

Linberg, K. R. (1999). Software developer perceptions about software project failure: A case study.
Journal of Systems and Software, 49(2). https://doi.org/10.1016/S0164-1212(99)00094-1

Lu, J., Yu, C. S., Liu, C., & Yao, J. E. (2003). Technology acceptance model for wireless Internet.
Internet Research, 13(3). https://doi.org/10.1108/10662240310478222

144
Maceachern, S. J., & Forkert, N. D. (2021). Machine learning for precision medicine. In Genome
(Vol. 64, Issue 4). https://doi.org/10.1139/gen-2020-0131

Magaña Martínez, D., & Fernandez-Rodriguez, J. C. (2015). Artificial Intelligence Applied to Project
Success: A Literature Review. International Journal of Interactive Multimedia and Artificial
Intelligence, 3(5). https://doi.org/10.9781/ijimai.2015.3510

Mahajan, G. (2021). Big Data and Predictive Analytics Drive Innovations in Construction Industry.
International Journal of Innovative Science & Technology, 8.

Mahdavinejad, M. S., Rezvan, M., Barekatain, M., Adibi, P., Barnaghi, P., & Sheth, A. P. (2018).
Machine learning for internet of things data analysis: a survey. In Digital Communications and
Networks (Vol. 4, Issue 3). https://doi.org/10.1016/j.dcan.2017.10.002

Mahdi, M. N., Zabil, M. H. M., Ahmad, A. R., Ismail, R., Yusoff, Y., Cheng, L. K., Mohd Azmi, M.
S., Natiq, H., & Naidu, H. H. (2021). Software project management using machine learning
technique-a review. Applied Sciences (Switzerland), 11(11).
https://doi.org/10.3390/app11115183

Malik, H., Afthanorhan, A., Amirah, N. A., & Fatema, N. (2021). Machine learning approach for
targeting and recommending a product for project management. Mathematics, 9(16).
https://doi.org/10.3390/math9161958

Marsland, S. (2014). Machine learning: An algorithmic perspective. In Machine Learning: An

Algorithmic Perspective, Second Edition. https://doi.org/10.1201/b17476

Matthies, C., & Hesse, G. (2019). Towards using data to inform decisions in agile software
development: Views of available data. ICSOFT 2019 - Proceedings of the 14th International
Conference on Software Technologies. https://doi.org/10.5220/0007967905520559

McGovern, A., Elmore, K. L., Gagne, D. J., Haupt, S. E., Karstens, C. D., Lagerquist, R., Smith, T.,
& Williams, J. K. (2017). Using artificial intelligence to improve real-time decision-making for
high-impact weather. Bulletin of the American Meteorological Society, 98(10).
https://doi.org/10.1175/BAMS-D-16-0123.1

Miklosik, A., Kuchta, M., Evans, N., & Zak, S. (2019). Towards the Adoption of Machine Learning-
Based Analytical Tools in Digital Marketing. IEEE Access, 7.
https://doi.org/10.1109/ACCESS.2019.2924425

Missonier, S., & Loufrani-Fedida, S. (2014). Stakeholder analysis and engagement in projects: From
stakeholder relational perspective to stakeholder relational ontology. International Journal of
Project Management. https://doi.org/10.1016/j.ijproman.2014.02.010

145
Mittelstadt, B. D., Allo, P., Taddeo, M., Wachter, S., & Floridi, L. (2016). The ethics of algorithms:
Mapping the debate. Big Data and Society, 3(2), 1–21.
https://doi.org/10.1177/2053951716679679

Moe, N. B., Aurum, A., & Dybå, T. (2012). Challenges of shared decision-making: A multiple case
study of agile software development. Information and Software Technology, 54(8).
https://doi.org/10.1016/j.infsof.2011.11.006

Mohammed, A. B. (2022). Applying BIM to achieve sustainability throughout a building life cycle
towards a sustainable BIM model. International Journal of Construction Management, 22(2).
https://doi.org/10.1080/15623599.2019.1615755

Mohammed, M., Khan, M. B., & Bashie, E. B. M. (2016). Machine learning: Algorithms and
applications. In Machine Learning: Algorithms and Applications.
https://doi.org/10.1201/9781315371658

Mohan, M., & Varghese, S. (2008). IRJET-Artificial Intelligence Enabled Safety for Construction
Sites Artificial Intelligence Enabled Safety for Construction Sites. International Research
Journal of Engineering and Technology.

Morley, J., Kinsey, L., Elhalal, A., Garcia, F., Ziosi, M., & Floridi, L. (2021). Operationalising AI
ethics: barriers, enablers and next steps. AI and Society. https://doi.org/10.1007/s00146-021-
01308-8

Munir, M. (2019). How Artificial Intelligence Can Help Project Managers. Global Journal of
Management and Business Research: Administration and Management, 19(4).

Naeini, F. H., & Krishnan, B. (2012). Usage pattern, perceived usefulness and ease of use of computer
games among malaysian elementary school students. Research Journal of Applied Sciences,
Engineering and Technology, 4(23).

Nasir, M. H. N. M., & Sahibuddin, S. (2011). Addressing a critical success factor for software
projects: A multi-round delphi study of TSP. International Journal of Physical Sciences, 6(5).

Obikunle, O. (2002). Dealing With Cultural Diversity in Project Management : A Dilemma in

Communication. Project Management Institute Annual Seminar and Symposium, San Antonio,
Texas, USA.

Pan, Y., & Zhang, L. (2021). Roles of artificial intelligence in construction engineering and
management: A critical review and future trends. In Automation in Construction (Vol. 122).
https://doi.org/10.1016/j.autcon.2020.103517

Panikkar, R., Saleh, T., Szybowski, M., & Whiteman, R. (2021). Operationalizing machine learning
in processes. McKinsey & Company, September.

146
Patten, M. L., & Patten, M. L. (2018). Qualitative Research Design. In Understanding Research
Methods. https://doi.org/10.4324/9781315213033-51

Pham, D. T., & Afify, A. A. (2005). Machine-learning techniques and their applications in
manufacturing. Proceedings of the Institution of Mechanical Engineers, Part B: Journal of
Engineering Manufacture, 219(5). https://doi.org/10.1243/095440505X32274

Pillai, M., Adapa, K., Das, S. K., Mazur, L., Dooley, J., Marks, L. B., Thompson, R. F., & Chera, B.
S. (2019). Using Artificial Intelligence to Improve the Quality and Safety of Radiation Therapy.
Journal of the American College of Radiology, 16(9). https://doi.org/10.1016/j.jacr.2019.06.001

Prasad, K. S. N., & Vijaya Saradhi, M. v. (2019). Comprehensive project management framework
using machine learning. International Journal of Recent Technology and Engineering, 8(2
Special Issue 3). https://doi.org/10.35940/ijrte.B1256.0782S319

Prifti, V. (2022). Optimizing Project Management using Artificial Intelligence. European Journal of
Formal Sciences and Engineering, 5(1), 29–37.

Procaccino, J. D., Verner, J. M., Darter, M. E., & Amadio, W. J. (2005). Toward predicting software
development success from the perspective of practitioners: An exploratory Bayesian model.
Journal of Information Technology, 20(3). https://doi.org/10.1057/palgrave.jit.2000044

Project Management Institute. (2017). PMBOK Guide - 6th Edition. In Project Management Institute
(Issue 2). PMI.

Provost, F., & Fawcett, T. (2013). Data Science and its Relationship to Big Data and Data-Driven
Decision Making. Big Data. https://doi.org/10.1089/big.2013.1508

Quazi, S. (2022). Artificial intelligence and machine learning in precision and genomic medicine. In
Medical Oncology (Vol. 39, Issue 8). https://doi.org/10.1007/s12032-022-01711-1

Rahimi, B., Nadri, H., Afshar, H. L., & Timpka, T. (2018). A systematic review of the technology
acceptance model in health informatics. In Applied Clinical Informatics (Vol. 9, Issue 3).
https://doi.org/10.1055/s-0038-1668091

Raval, R., & Rathod, H. (2014). Improvements in Agile Model using Hybrid Theory for Software
Development in Software Engineering. International Journal of Computer Applications, 90(16).
https://doi.org/10.5120/15806-4677

Robert, C. (2014). Machine Learning, a Probabilistic Perspective .

CHANCE, 27(2). https://doi.org/10.1080/09332480.2014.914768

Rožman, M., Oreški, D., & Tominc, P. (2023). Artificial-Intelligence-Supported Reduction of

Employees’ Workload to Increase the Company’s Performance in Today’s VUCA Environment.
Sustainability, 15(6). https://doi.org/10.3390/su15065019

147
Rudolf, G. (2023). IMPLEMENTATION BARRIERS OF ARTIFICIAL INTELLIGENCE IN
COMPANIES. Proceedings of FEB Zagreb 14 Th International Odyssey: Conference on
Economics and Business.

Sang, L., Yu, M., Lin, H., Zhang, Z., & Jin, R. (2021). Big data, technology capability and
construction project quality: a cross-level investigation. Engineering, Construction and
Architectural Management, 28(3), 706–727. https://doi.org/10.1108/ECAM-02-2020-0135

Sarker, I. H. (2021). Machine Learning: Algorithms, Real-World Applications and Research

Directions. In SN Computer Science (Vol. 2, Issue 3). https://doi.org/10.1007/s42979-021-
00592-x

Sarker, I. H., Abushark, Y. B., Alsolami, F., & Khan, A. I. (2020). IntruDTree: A machine learning
based cyber security intrusion detection model. Symmetry.
https://doi.org/10.3390/SYM12050754
Sarker, I. H., Abushark, Y. B., & Khan, A. I. (2020). ContextPCA: Predicting context -aware
smartphone apps usage based on machine learning techniques. Symmetry, 12(4).
https://doi.org/10.3390/SYM12040499

Sarker, I. H., Hoque, M. M., Uddin, M. K., & Alsanoosy, T. (2021). Mobile Data Science and
Intelligent Apps: Concepts, AI-Based Modeling and Research Directions. Mobile Networks and
Applications, 26(1). https://doi.org/10.1007/s11036-020-01650-z

Sarker, I. H., Kayes, A. S. M., Badsha, S., Alqahtani, H., Watters, P., & Ng, A. (2020). Cybersecurity
data science: an overview from machine learning perspective. Journal of Big Data.
https://doi.org/10.1186/s40537-020-00318-5

Sarker, I. H., Kayes, A. S. M., & Watters, P. (2019). Effectiveness analysis of machine learning
classification models for predicting personalized context-aware smartphone usage. Journal of
Big Data. https://doi.org/10.1186/s40537-019-0219-y

Saunders, M. N. K. (2017). Choosing Research Participants. In Qualitative Organizational Research:

Core Methods and Current Challenges. https://doi.org/10.4135/9781526435620.n3

Saunders, M. N. K., & Townsend, K. (2016). Reporting and Justifying the Number of Interview
Participants in Organization and Workplace Research. British Journal of Management, 27(4).
https://doi.org/10.1111/1467-8551.12182

Schwaber, K., & Sutherland, J. (2011). The Scrum Guide - The Definitive Guide to Scrum: The Rules
of the Game. In Scrum. org, October (Vol. 2, Issue October).

Schwarz, J., Sandoval-Wong, A., & Sanchez, P. M. (2015). Implementation of artificial intelligence
into risk management decision-making processes in construction projects.

148
Shang, G., Low, S. P., & Lim, X. Y. V. (2023). Prospects, drivers of and barriers to artificial
intelligence adoption in project management. Built Environment Project and Asset Management.
https://doi.org/10.1108/BEPAM-12-2022-0195

Sheffield, J., & Lemétayer, J. (2013). Factors associated with the software development agility of
successful projects. International Journal of Project Management, 31(3).
https://doi.org/10.1016/j.ijproman.2012.09.011

Shenton, A. K. (2004). Strategies for ensuring trustworthiness in qualitative research projects.

Education for Information, 22(2). https://doi.org/10.3233/EFI-2004-22201

Smith, A., & Fressoli, M. (2021). Post-automation. Futures, 132.

https://doi.org/10.1016/j.futures.2021.102778

Sravanthi, J., Sobti, R., Semwal, A., Shravan, M., Al-Hilali, A. A., & Bader Alazzam, M. (2023). AI-
Assisted Resource Allocation in Project Management. 2023 3rd International Conference on
Advance Computing and Innovative Technologies in Engineering, ICACITE 2023.
https://doi.org/10.1109/ICACITE57410.2023.10182760

Standish Group. (2015). Chaos Report 2015.

Steels, L., & de Mantaras, R. L. (2018). The Barcelona declaration for the proper development and
usage of artificial intelligence in Europe. AI Communications, 31(6).
https://doi.org/10.3233/AIC-180607

Stewart, M. (2019). The Limitations of Machine Learning - Towards Data Science. In Medium -
Towards Data Science.

Su, L.C., Wu, X., Zhang, X., & Huang, X. (2021). Smart performance-based design for building fire
safety: Prediction of smoke motion via AI. Journal of Building Engineering, 43.
https://doi.org/10.1016/j.jobe.2021.102529

Sundaram, S., & Zeid, A. (2023). Artificial Intelligence-Based Smart Quality Inspection for
Manufacturing. Micromachines, 14(3). https://doi.org/10.3390/mi14030570

Surendran, P. (2012). Technology Acceptance Model : A Survey of Literature. International Journal

of Business and Social Research, 2(4).

Tambe, P., Cappelli, P., & Yakubovich, V. (2019). Artificial intelligence in human resources
management: Challenges and A path forward. California Management Review, 61(4).
https://doi.org/10.1177/0008125619867910

149
Tantiponganant, P., & Laksitamas, P. (2014). An analysis of the technology acceptance model in
understanding students’ behavioral intention to use university’s social media. Proceedings - 2014
IIAI 3rd International Conference on Advanced Applied Informatics, IIAI-AAI 2014.
https://doi.org/10.1109/IIAI-AAI.2014.14

Tiwari, T., Tiwari, T., & Tiwari, S. (2018). How Artificial Intelligence, Machine Learning and Deep
Learning are Radically Different? International Journal of Advanced Research in Computer
Science and Software Engineering, 8(2). https://doi.org/10.23956/ijarcsse.v8i2.569

Torrecilla-García, J. A., Pardo-Ferreira, M. C., & Rubio-Romero, J. C. (2021). Overall introduction to

the framework of BIM-based digital twinning in decision-making in safety management in
building construction industry. Direccion y Organizacion, 74, 31–38.
https://doi.org/10.37610/DYO.V0I74.600

Tran, S. V. T., Nguyen, T. L., & Park, C. (2021). A BIM Integrated Hazardous Zone Registration
Using Image Stitching. Proceedings of the International Symposium on Automation and Robotics
in Construction, 2021-November. https://doi.org/10.22260/isarc2021/0026

Ugwudike, P. (2022). AI audits for assessing design logics and building ethical systems: the case of
predictive policing algorithms. AI and Ethics, 2(1). https://doi.org/10.1007/s43681-021-00117-5

Ulrich, P., Frank, V., & Kratt, M. (2021). Adoption of artificial intelligence technologies in German
SMEs — Results from an empirical study. https://doi.org/10.22495/cgsetpt13

Uysal, M. P. (2021). Machine Learning and Data Science Project Management From an Agile
Perspective. https://doi.org/10.4018/978-1-7998-7872-8.ch005

Vanhoucke, M., Coelho, J., & Batselier, J. (2016). An overview of project data for integrated project
management and control. In Journal of Modern Project Management (Vol. 3, Issue 3).

Vuković, M., Pivac, S., & Kundid, D. (2019). Technology Acceptance Model for the Internet Banking
Acceptance in Split. Business Systems Research, 10(2). https://doi.org/10.2478/bsrj-2019-022

Waltersmann, L., Kiemel, S., Stuhlsatz, J., Sauer, A., & Miehe, R. (2021). Artificial intelligence
applications for increasing resource efficiency in manufacturing companies—A comprehensive
review. In Sustainability (Switzerland) (Vol. 13, Issue 12). https://doi.org/10.3390/su13126689

Wang, Q. (2019). How to apply AI technology in Project Management. PM World Journal How to
Apply AI Technology in Project Management, VIII(III).

Wang, S. Q., Dulaimi, M. F., & Aguria, M. Y. (2004). Risk management framework for construction
projects in developing countries. Construction Management and Economics, 22(3).
https://doi.org/10.1080/0144619032000124689

150
Waymel, Q., Badr, S., Demondion, X., Cotten, A., & Jacques, T. (2019). Impact of the rise of
artificial intelligence in radiology: What do radiologists think? Diagnostic and Interventional
Imaging, 100(6). https://doi.org/10.1016/j.diii.2019.03.015

Wei, W., & Rana, M. E. (2019). Software project schedule management using machine learning &
data mining. International Journal of Scientific and Technology Research, 8(9).

Witten, I. H., Frank, E., Hall, M. A., & Pal, C. J. (2016). Data Mining: Practical Machine Learning
Tools and Techniques. In Data Mining: Practical Machine Learning Tools and Techniques.
https://doi.org/10.1016/c2009-0-19715-5

Wu, D. D., Chen, S. H., & Olson, D. L. (2014). Business intelligence in risk management: Some
recent progresses. Information Sciences, 256. https://doi.org/10.1016/j.ins.2013.10.008

Zainal, Z. (2007). Case study as a research method. Jurnal Kemanusiaan Bil, 9.

Zannier, C., Chiasson, M., & Maurer, F. (2007). A model of design decision making based on
empirical results of interviews with software designers. Information and Software Technology,
49(6). https://doi.org/10.1016/j.infsof.2007.02.010

Zhao, X., Hwang, B. G., & Gao, Y. (2016). A fuzzy synthetic evaluation approach for risk
assessment: A case of Singapore’s green projects. Journal of Cleaner Production, 115.
https://doi.org/10.1016/j.jclepro.2015.11.042

Zhou, Y., Wang, Y., Chen, H., Xu, Y., Luo, Y., Deng, Y., Zhang, J., & Shao, A. (2021). Immuno-
oncology: are TAM receptors in glioblastoma friends or foes? In Cell Communication and
Signaling (Vol. 19, Issue 1). https://doi.org/10.1186/s12964-020-00694-8

Zulch, B. (2014). Leadership Communication in Project Management. Procedia - Social and

Behavioral Sciences. https://doi.org/10.1016/j.sbspro.2014.03.021

151
 APPENDIX A. INTERVIEW GUIDE

Prepare the recording device and schedule the interviews based on convenience.

Introduction:

Welcome, and thank you for your participation in this interview. The focus of the

interview of gaining insight from you in your industry regarding the adoption and use of

Artificial Intelligence, hereafter referred to as AI, and Machine Learning, hereafter referred to as

ML, in project management.

Ask the candidate how they are doing and provide a brief on what the research aims to do

as well as its progress and inform them that the interview is being recorded (Audio only).

Next, request them to introduce themselves, names withheld, with a focus on industry,

experience, and position.

Explore the set of questions focused on understanding the experience and perception of

the individuals with project management and the use of AI/ML.

• Please describe the type of projects and the project management function in your

organization.

• Is Artificial Intelligence used in any capacity in project management? If yes,

please explain its purpose and the specific activity the AI focuses on.

• Based on your experience, do you believe AI is useful for project management?

Which areas do you think it has helped?

Continue the focus on the complexity of the system, benefits, barriers, and its importance.

• AI is a complex system involving many different tools and systems that could be

used. How easy is it to identify the best fit for your requirements? What are some

challenges that might be faced in its use?

152
 • Once trained, do you believe the system is effective and easy to use?

• Do you believe that Machine Learning systems are critical for project

management and its success?

• Many studies show ML to be capable in literature to improve decision-making.

What is your view on it?

• What do you believe is the benefit of ML based on your experience?

• What do you think is the primary barrier to its adoption?

Next, focus on questions that are focused on understanding how things could be changed

and what would be the future of technology.

• What are some areas where improvement is required for improved acceptance?

• Based on your experience, do you feel there is a more considerable reluctance to

adopt ML?

• Do you feel the technology would be accepted and integrated into PM practices

across your industry?

Ask the participants if they are comfortable, and thank them for their participation.

Enquire as to if they would like to know the final result, and if yes, inform them that they will be

informed after completion. Once again, thank them and stop the recording.

153
 ProQuest Number: 30991003

INFORMATION TO ALL USERS

The quality and completeness of this reproduction is dependent on the quality
and completeness of the copy made available to ProQuest.

Distributed by ProQuest LLC ( 2024 ).

Copyright of the Dissertation is held by the Author unless otherwise noted.

This work may be used in accordance with the terms of the Creative Commons license
or other rights statement, as indicated in the copyright statement or in the metadata
associated with this work. Unless otherwise specified in the copyright statement
or the metadata, all rights are reserved by the copyright holder.

This work is protected against unauthorized copying under Title 17,

United States Code and other applicable copyright laws.

Microform Edition where available © ProQuest LLC. No reproduction or digitization

of the Microform Edition is authorized without permission of ProQuest LLC.

ProQuest LLC
789 East Eisenhower Parkway
P.O. Box 1346
Ann Arbor, MI 48106 - 1346 USA