‘Lean 4.

0’:
How can digital technologies support lean practices?

Fabiana Dafne Cifone (fabianadafne.cifone@polimi.it)

Dept. of Management, Economics and Industrial Engineering, Politecnico di Milano

Kai Hoberg
Kühne Logistics University

Matthias Holweg
Saïd Business School, University of Oxford

Alberto Portioli Staudacher

Dept. of Management, Economics and Industrial Engineering, Politecnico di Milano

Abstract
The digitalisation of business offers great possibilities to create new products and
processes, as well as improve existing ones. In this paper we seek to understand how
digital technologies can support process improvement in a manufacturing context.
Several studies have proposed synergies among digitalization and lean at a conceptual
level, yet so far we lack any empirical proof. To this effect, we present exploratory
quantitative research aimed at explaining how digital technologies can support lean
practices, improving our understanding how to harness the potential of digitalisation in
operational improvement. We conclude with areas for further research.

Keywords: Lean, process improvement, digitalization, manufacturing.

Introduction
Digital technologies are powerful innovations that have rightfully captured the
imagination of manufacturing managers. It has been frequently stated that digital
technologies can support or enhance lean practices. Here, technologies often associated
with ‘Industry 4.0’ are cited, and we will refer to such digital enhancements of lean
practices as ‘Lean 4.0’. For example, according to Wagner et al. (2017), digitalization
will not only support lean practices, but it will also enlarge their scope. Even on the
1
performance point of view, researchers reached the consensus on the operational
improvements lean practices yield thanks to digitalization (i.e. Davies et al., 2017;
Kolberg and Zühlke, 2015; Tortorella et al., 2018). However, beyond these early studies,
to the best of authors’ knowledge, conclusive proof of this claim is still outstanding.
It is easy to conceive areas where they directly support lean practices, yet so far much of
this discourse is based on conjecture. With the underlying study we seek to provide a
rigorous identification of the mechanisms explaining how digital technologies can
support lean practices, and thus contribute to our theoretical understanding of the true
impact of ‘Industry 4.0’ technologies, and to support managerial decision in making the
business case for their adoption.
In this paper, we present a structured literature review of lean practices as well as
digital technologies of relevance to operations management. We show findings of an
initial survey that demonstrated the lack of maturity of digital technology adoption in
practice and highlighted the needs for further qualitative research. We conclude with
comments on preliminary results and presenting our next steps.

Theoretical background
Lean: wastes and practices
The concept of ‘lean production’ has been widely researched and discussed in the
operations management literature (c.f. Holweg, 2007, Fujimoto, 1999, and many others).
At the very heart of lean stands the concept of waste reduction, namely to improve a
process by reducing non-value activities herein. Taiichi Ohno coined the original seven
wastes (or ‘muda’) in manufacturing (often abbreviated as TIMWOOD – Transportation,
(excess) Inventory, Motion, Waiting, Overproduction, Overprocessing and Defects),
which later have been expanded to also include ‘Skills’, or wasted human talent and ideas.
Muda, together with ‘mura’ (unevenness) and ‘muri’ (overburden), provides the original,
and still the most succinct, way how to conceptualise lean (Bicheno and Holweg, 2016).
It is for that reason that we adopt the seven (eight) wastes for our study, coupled with
the lean practices that have been built upon them to form an integrated management
system (Shah and Ward, 2003; Womack, Jones and Roos, 1990). These practices work
synergistically to achieve the main goals of lean production, which are related to the
creation of a streamlined, high quality system that produces finished products at the pace
of customer demand with little or no waste (Shah and Ward, 2003). For their nature, lean
practices are applicable to the entire company (Ruiz-Benítez, López and Real, 2018).
However, in the scientific community, there is no a unique classification of practices or
consensus on which is the complete set of lean practices. Different authors, in fact, refers
to them providing their own classification or list of practices. For this reason, and for the
scope of this study, a systematic literature review has carried out, with the aim to define
a comprehensive list of lean practices.
We adopted an ad-hoc keywords strategy to article title/abstract and keywords on
Scopus database. Keywords selected are based on the term “lean” combined with the
Boolean operator AND to terms referring to “practice” or “bundle”. Moreover, we

2
adopted five inclusion criteria. Firstly, we reviewed only contributions in English.
Secondly, we considered only journal paper belonging to high-ranked journal (Q1
according to Scimago ranking), discarding conference contributions, books chapter or
journal paper published in a medium-to-low ranked journal (according to Scimago
ranking). Thirdly, we included only contributions with a proper and specified
review/classification of practices. Fourthly, we restricted our focus to manufacturing
industry, discarding then all contributions clearly referred to service industries.
Eventually, we considered contributions from 2000 on.
The review process identified 528 eligible studies from all the keywords. We filtered
the studies by scanning their titles and abstracts, removing duplicates and selecting those
consistent with the aforementioned inclusion criteria, resulting in 179 articles. Full-texts
were assessed with the same criteria, to discard out-of-scope documents, resulting in 81
articles. 140 practices have been identified, with more than 75% of them cited less than
10 times, highlighting the lack of structured and comprehensive list of practices.
Following the final selected sample of practises considered, having each a number of
citations at least equal to 10.

Table 1 - Lean practices

Lean practice Number of citations
JIT(Just in Time/ Continuous flow production) 66
Pull system (Kanban) 56
Quick changeover techniques and reduction of setup time (SMED) 53
TPM (Total productive maintenance) 47
Continuous improvement programs (Kaizen) 39
TQM (Total quality management/Zero defects) 37
Supplier involvement and developement (feedback and partnership) 34
Production smoothing (bottleneck removal, Heijunka) 31
Cross-functional work force 31
Cellular manufacturing 30
VSM (Value Stream Mapping) 28
5S 27
Work standardization (SOPs stand. operating procedures) 26
Error proofing (Poka-Yoke) 26
Lot size reductions 23
VLPM (Visual Performance Measures/Visual control) 23
Customer involvement and partnership (feedback) 22
Statistical Process Control (SPC) 17
Employees' involvement (suggestion schemes) 16
Autonomation (Jidoka) 16
Information sharing 15
Lean Management Training 15
Elimination of waste 14
Shop floor organization and safety 14
Small group problem solving 14
Preventive maintenance 13
Low inventory 13
HRM (Human Resources Management) 12
Top management leadership for quality 11
Reduced number of suppliers 10
Takt time definition 10

The resulted panel is made by 31 lean practices, with various degree of frequency. JIT
and Pull systems are referred to most frequently, while Smaller number of suppliers and
Takt time definition are of less interest in the reviewed literature.
As mentioned before, according to its definition, lean strives to minimize general
understanding of waste (Womack, Jones and Roos, 1990). It is essential hence to
3
understand the role played by lean practices on the 8 wastes. Here following a table
summarizing which lean practice acts on specific waste. As result of reviewing literature,
some practices result to have a more horizontal impact on wastes compared to others. As
example, Kaizen events affect positively all 8 wastes, while the takt time definition results
to impact only on waiting and inventories.

Table 2 - Lean practices and 8 wastes

Lean practice Transportation Inventories Motion Waiting Overproduction Overoprocessing Defects Skills
JIT(Just in Time/ Continuous flow production) x x x x
Pull system (Kanban) x x x
Quick changeover techniques and reduction of setup time (SMED) x x x
TPM (Total productive maintenance) x x
Continuous improvement programs (Kaizen) x x x x x x x x
TQM (Total quality management/Zero defects) x x x
Supplier involvement and developement (feedback and partnership) x x
Production smoothing (bottleneck removal, Heijunka) x x
Cross-functional work force x x
Cellular manufacturing x x
VSM (Value Stream Mapping) x x x
5S x x
Work standardization (SOPs stand. operating procedures) x x x
Error proofing (Poka-Yoke) x x
Lot size reductions x x
VLPM (Visual Performance Measures/Visual control) x x x x
Customer involvement and partnership (feedback) x x x
Statistical Process Control (SPC) x x x x
Employees' involvement (suggestion schemes) x x x x x x x x
Autonomation (Jidoka) x x x
Information sharing x x x x x x x x
Lean Management Training x x x x x x x x
Elimination of waste x x x x x x x x
Shop floor organization and safety x x
Small group problem solving x
Preventive maintenance x x
Low inventory x x
HRM (Human Resources Management) x x
Top management leadership for quality x x x x x x x
Reduced number of suppliers x x
Takt time definition x x

Digital technologies
The availability of low-cost sensors, increases of computing power and high-speed
internet connectivity are some of the enablers of massive advances in technologies for
operations and supply chain management. Companies have always used new technologies
to advance their process. The shipping container is probably the most successful example
of a technical revolution that not only significantly improved supply chain processes but
also shaped global trade flows in the long term (Cooper and Levinson, 2010).
On the other hand, there are also technologies like Radio Frequency Identification
(RFID) that have triggered high expectations for process improvement in retailing
operations, but have so far only been able to partially fulfil (Gaukler and Seifert, 2007).
Ultimately, new technologies can provide benefits in two fundamentally different ways,
i.e. either by increasing efficiency or by increasing revenues – many technologies aim to
achieve both. As we are interested in the technology benefits that aim at waste reduction,
we carefully screened for relevant technologies in this context. Since new technology
developments are currently observed every day, we focused our literature analysis on
practitioner articles as well as white papers and reports issued by large technology firms

4
and big consultancies. While the mentioned technologies differ slightly and the
terminology is also varying a comprehensive picture emerged that put six technology
clusters in the focus of our interests. Table 3 provides a summary of the technologies that
we identified as potentially most relevant for process improvement.

Table 3 - Summary of technologies

Technology Description Examples
IoT solutions Sensors, cameras and smart devices that process and Process control sensors,
share gathered data using internet connectivity. environmental monitoring,
cameras, smart replenishment
solutions.
Virtual and Interactive experience where real world objects are Smart glass, holo-lens, virtual
augmented reality either represented completely “virtual” or twins.
"augmented" by computer-generated perceptual
information.
Advances Data science tools for improved decision making, Predictive Analytics, Machine
analytics e.g., by gaining deeper insights, making predictions, Learning, Deep Learning,
or generating recommendations. Support Vector Machines.
Autonomous Solutions allowing semi- and fully autonomous Platooning trucks, autonomous
vehicles transportation, ranging from long-distance to short- trucks, drones, self driving
distance deliveries. delivery vehicles.
Robotics Physical robotic systems used across all supply Robotic mobile fulfillment
chain processes within enclosed environments. systems, picking robots, industrial
robots, cobots.
Digital Integrated, computer-based system manufacturing Digital printing, 3D printing, CNC
manufacturing comprised of simulation, 3D visualization, analytics milling, Stereolithography.
and collaboration tools to create product and
manufacturing process.

‘Lean 4.0’
The integration between lean and Industry 4.0 (Lean 4.0) is currently highly discussed in
literature (Buer, Strandhagen and Chan, 2018). Even if academia seems to agree on the
potentials of the aforementioned integration, there are different points of view depending
on the author explaining the subject. Two main perspectives can be drafted in the
academia. Some authors describe lean as a basis for the implementation of Industry 4.0
(Hambach, Kümmel and Metternich, 2017). Indeed, since lean practices are aimed at
wastes reduction along the process (Womack and Jones, 2003), having a streamlined and
under control process represents the prerequisite for any process digitalization (Buer,
Strandhagen and Chan, 2018). Other studies have significantly confirmed that companies
5
with a higher associated level of lean implementation benefit the most in embracing
Industry 4.0 and in grasping its potentials (Hoellthaler, Braunreuther and Reinhart, 2018).
Other researchers refer instead to Industry 4.0 as a completion of lean (Kolberg and
Zühlke, 2015), that was declared as limited by some studies. Market requirements are
nowadays more complex and customers demand for highly personalized products may
hinder lean to be still effective. Lean could not only be able to keep up with the pace of
personalization using the same tools used since the second half of the 20th century with
no technological advancements supporting those tools (Sanders, Elangeswaran and
Wulfsberg, 2016). In this sense, Industry 4.0 represents the mean lean can exploit to face
new trends in the manufacturing world, preserving its process’ robustness.
The strong interest for the topic from the academia is evident, however due to the infancy
of the Industry 4.0 topic, it is still difficult to assess the effect of Lean 4.0. To the best of
authors’ knowledge, available scientific studies are mostly focused on theoretical
research, and hence conclusive proof for Lean 4.0 potentials is still outstanding.
This research is an attempt to study the mechanisms explaining how digital
technologies can enhance lean, and to assess the impact of ‘Lean 4.0’ on operational
performance. The scope is limited to industrial operations management, including
manufacturing, logistics and supply chain operations, since it is the traditional application
space for lean.

Exploratory survey
In this section we will we will present our exploratory survey carried out in the European
manufacturing sector aimed at deepen the relationship between Lean and digital
technologies adoption, as well as the effects on operational performance. Firstly, we will
describe the design of the survey and its instrument, as well as its implementation.
Secondly, we will present first preliminary results of our quantitative study.

Design
A survey methodology has been selected as the most suitable one among quantitative
methodologies.
The questionnaire is made by 19 questions, grouped into six main clusters: (i)
companies’ profile; (ii) contextual factors; (iii) lean practice implementation; (iv) digital
technology implementation; (v) the effect of ‘Lean 4.0’; (vi) operational performance.
The constructs related to lean practices and digital technologies used in the questionnaire
are based on the reviewed literature both on lean practices and on technologies. For what
regards operational performance, both questions and response interval come from the
study of Shah and Ward (2003). Moreover, set-up time and inventories have been
included as additional items due to their relevance for the lean approach. The response
interval for these items is built following the same structure suggested by Shah and Ward
(2003).
The sample selected for this exploratory survey is limited to the European
manufacturing sector, with plants as unit of analysis. Respondents are required to be

6
experienced in lean, therefore green belt and black belt experts have been targeted. These
criteria led to a non-random choice of companies for the survey, a strategy used also in
the study of Shah and Ward (2003). In order to define the size of the sample, it was
considered that 15% is the average rate in management surveys and 100 is the minimum
threshold of responses to perform a significant statistical analysis (Hair et al., 2007). On
the base of these numbers, sample considered valid consisted in about 1200 experts.

Implementation
The questionnaire instrument has been tested with a sample of 56 lean experts to verify
and to validate the constructs and the questions. Testing phase started in May 2018 and
lasted 5 months. The final and improved online survey, through Google Module, has been
submitted from November 2018 to February 2019 to more than 1200 lean experts.
Questionnaire submission exploited three different waves: first wave addressed 700
people, second wave 300 people, and third wave about 200 people. Non-response bias
has been managed through two actions: first, while sending the survey it was specified
that the questionnaire lasts at least 10 minutes; secondly, the questionnaire was sent three
times with a time window of about one week.
The final number of completed surveys was 162 corresponding to a response rate of
approximately 13,32%. The validity of answers was verified excluding answers coming
from experts involved in service sector (56 answers) and cleaning the dataset from clearly
random responses (1 answer). The final dataset is composed by 105 responses, referring
to 88 different companies. In addition to all the answers provided, final dataset
comprehends two additional values for each expert/plant: lean level (LL) and digital level
(DL). According to the model developed by Soriano-Meier and Forrester (2002), LL can
be defined starting from a self-evaluation on the implementation of several lean practices.
The final LL comes from the computation of the average value of self-evaluation values.
The same reasoning of LL has been used for computing DL, exploiting the six digital
tools considered in the survey.

Results
In terms of production technology, responses are equally distributed among process and
discrete manufacturing. Moreover, several industrial sectors are represented in the
sample. Most of the respondents report their plant age as higher than 20 years while only
about 7% of the plants have been built in the last decade (Figure 1 - A). About the plant
size, according to the European classification, the sample is representative of medium-
large plants (Figure 1 - B). Only about the 6,7% of the respondents is indeed in small
plants (i.e. less than 50 employees).
Eventually, it is interesting to understand how respondents are spread out on different
departments. It is indeed necessary to stress that all the respondents are lean experts,
endowed with green or black belt certification, and are working in different departments
within companies. 31,4% of the respondents are involved in the Production and
Maintenance departments. A remarkable percentage of experts, 27,6%, belongs instead

7
to the so-called Kaizen Promotion Office, that is usually a support-function in charge of
continuous improvement activities of the company. Quality and Supply chain functions
are also represented in the sample.

Figure 1 - Age plants (A) and Size plants (B)

Data gathered from the online survey have been using SPSS, a software of data
mining and data statistical analysis.
In order to study the dependence of DL on LL, a box plot analysis was carried out to
better understand how the data are distributed. As displayed in Figure 2 - A, not only data
are not normally distributed but also DL shows a lower variability than LL. It is
interesting to note that there are three outliners at the top of the graph: these three plants
are classified as “old” and “large” and are associated with a very high LL. Same data of
DL and LL are plotted in a scatter plot to have a graphical representation of the relation
between the two variables (Figure 2 - B).
The second quadrant of the matrix presents a high concentration of observation,
meaning that most of the plants having a low DL report a high LL. Moreover, not only
the fourth quadrant is empty - there are no plants with high DL and low LL, but also DL
is never higher than LL. This may confirm what is stated in the literature: highly
digitalized plants are also strongly implementing lean, making lean the prerequisite for
digital transformation (e.g. Buer, Strandhagen and Chan, 2018).

Figure 2 - DL and LL: Box plot (A) and scatter plot (B)

8
Despite this, too few plants registered an above average DL. Only about 10% of plants
are belonging to the first quadrant, registering high DL and high LL. This prevented us
to make meaningful statements on the possible digital mechanisms that affect lean.
Moreover, looking at operational performance results, the majority of plants is not at
the moment affected by any significant operational performance increase. A cluster
analysis of plants on operational performance highlights an interesting fact. Applying K-
means methods, we ended with 3 different clusters (k=3 defined by Ward distance
criterion) with plants not equally distributed among them. About 60% of plants indeed
belongs to cluster 3. If cluster 1 and cluster 2 registered remarkable improvement in one
or more operational performance, cluster 3 includes companies where Lean 4.0 did not
affect operational performance. As shown in Figure 3, it is interesting to note that plants
associated with high LL and high DL register a positive impact on operational
performance. This represents an initial preliminary result on the performance
improvement yielded by Lean 4.0. Unfortunately, further considerations cannot be draft
due to the low numbers of observations available with high LL and high DL.

Figure 3 - Clustering analysis on operational performance

Conclusion and next steps

Preliminary results coming from our exploratory survey showed a positive relationship
between the combined lean and digital adoption on operational performance. Low digital
maturity levels however prevent us from fully disentangling the effects as the majority of
plants still feature a low digital level. Thus, we will expand our research design to include
qualitative research to identify the potential mechanisms how digital can support lean that
practitioners foresee. Using a focus group-based study design (currently underway) we
seek to identify both potential variation of impact, as well as level of importance, of the
main digital technologies, as well as clustered mechanisms that vary across supply chain
activities. In other words, we want to understand which digital technologies have the
greatest impact on lean process improvement, and secondly, what the general
mechanisms are how digital technologies support lean. Based on our combined survey
and focus group findings we will provide a discourse on the how mechanisms identified
link to the wide theoretical landscape of lean improvements, and how this can inform
managerial practice to harness the potential that digital technologies offer to operations
management more generally.

9
References
Bicheno, J. and Holweg, M. (2016) ‘The Lean Toolbox, A handbook for lean transformation’, Production
and inventory control, systems and industrial engineering books. doi: 10.1016/j.sse.2009.12.024.
Buer, S. V., Strandhagen, J. O. and Chan, F. T. S. (2018) ‘The link between Industry 4.0 and lean
manufacturing: mapping current research and establishing a research agenda’, International Journal of
Production Research, pp. 1–17. doi: 10.1080/00207543.2018.1442945.
Cooper, R. N. and Levinson, M. (2010) ‘The Box: How the Shipping Container Made the World Smaller
and the World Economy Bigger’, Foreign Affairs. doi: 10.2307/20032089.
Davies, R., Coole, T. and Smith, A. (2017) ‘Review of Socio-technical Considerations to Ensure Successful
Implementation of Industry 4.0’, Procedia Manufacturing. Pp. 1288–1295. doi:
10.1016/j.promfg.2017.07.256.
Fujimoto, T. (1999), The evolution of manufacturing system at Toyota, Oxford University Press, New York.
Gaukler, G. M. and Seifert, R. W. (2007) ‘Applications of RFID in Supply Chains’, in Trends in Supply
Chain Design and Management. doi: 10.1007/978-1-84628-607-0_2.
Hair, J. F. et al. (2007) Multivariate Data Analysis (6th ed.), Pearson Prentice Hall, Upper Saddle River.
Hambach, J., Kümmel, K. and Metternich, J. (2017) ‘Development of a Digital Continuous Improvement
System for Production’, Procedia CIRP. Pp. 330–335. doi: 10.1016/j.procir.2017.03.086.
Hoellthaler, G., Braunreuther, S. and Reinhart, G. (2018) ‘Digital Lean Production-An Approach to Identify
Potentials for the Migration to a Digitalized Production System in SMEs from a Lean Perspective’, in
Procedia CIRP. doi: 10.1016/j.procir.2017.12.255.
Holweg, M. (2007) ‘The genealogy of lean production’, Journal of Operations Management, 25(2), pp.
420–437. doi: 10.1016/j.jom.2006.04.001.
Kolberg, D. and Zühlke, D. (2015) ‘Lean Automation enabled by Industry 4.0 Technologies’, IFAC-
PapersOnLine, 28(3), pp. 1870–1875. doi: 10.1016/j.ifacol.2015.06.359.
Ruiz-Benítez, R., López, C. and Real, J. C. (2018) ‘The lean and resilient management of the supply chain
and its impact on performance’, International Journal of Production Economics. doi:
10.1016/j.ijpe.2018.06.009.
Sanders, A., Elangeswaran, C. and Wulfsberg, J. (2016) ‘Industry 4.0 implies lean manufacturing: Research
activities in industry 4.0 function as enablers for lean manufacturing’, Journal of Industrial Engineering
and Management, 9(3), pp. 811–833. doi: 10.3926/jiem.1940.
Shah, R. and Ward, P. T. (2003) ‘Lean manufacturing: Context, practice bundles, and performance’,
Journal of Operations Management, 21(2), pp. 129–149. doi: 10.1016/S0272-6963(02)00108-0.
Soriano-Meier, H. and Forrester, P. L. (2002) ‘A model for evaluating the degree of leanness of
manufacturing firms’, Integrated Manufacturing Systems. doi: 10.1108/09576060210415437.
Tortorella, G. et al. (2018) ‘The mediating effect of employees’ involvement on the relationship between
Industry 4.0 and operational performance improvement’, Total Quality Management and Business
Excellence. Taylor & Francis, 0(0), pp. 1–15. doi: 10.1080/14783363.2018.1532789.
Wagner, T., Herrmann, C. and Thiede, S. (2017) ‘Industry 4.0 Impacts on Lean Production Systems’, in
Procedia CIRP. doi: 10.1016/j.procir.2017.02.041.
Womack, J. P. and Jones, D. (2003) ‘Lean Thinking − Banish waste and create wealth in your corporation’,
BMJ quality safety, 21(April), pp. 15–90. doi: 10.1136/bmjqs-2011-000173.
Womack, J. P., Jones, D. T. and Roos, D. (1990) ‘The Machine that Changed the World: The Story of Lean
Production’, Harper Perennial, New York.

10