THIRST FOR INNOVATION

ROUNDTABLE FINDINGS:
AI, UNIVERSAL INTELLIGENT
INFRASTRUCTURE MANAGEMENT
(UIIM), AND ALL THINGS
SUSTAINABILITY

DATE: JANUARY 2025

EDITION 1
AUTHOR(S): VARIOUS
AI, UIIM, AND ALL THINGS SUSTAINABILITY

CONTRIBUTORS

EXECUTIVE SUMMARY RiT Tech Leaders

• Jeff Safovich, CTO

This white paper provides insights from the Thirst for Innovation
(TFI) roundtable which was held on September 26, 2024. The • Mark Acton, Consultant
session brought together professionals from the technology
• Susan Anderton, Marketing Lead
and data center industries to discuss the impact of AI on
infrastructure and operations, the emerging methodology of • Antonio Suarez Garcia,
Global Product Manager
Universal Intelligent Infrastructure Management (UIIM), and the
challenges of sustainability reporting for EU compliance. Round Table September 2024 Delegates

The paper delves into the potential role of AI and machine Circa 30 delegates from a range
learning to optimize infrastructure and operations, particularly of operators and the data center
in relation to achieving desired energy efficiency levels for industry ecosystem. See Appendix 2.
operators and the opportunity for automated capabilities such as
predictive maintenance – all of which can improve the accuracy of
decision-making to meet compliance and commercial goals.

Ultimately, the paper highlights that, while there is a considerable

amount of uncertainty, embracing AI in relation to better
efficiencies in the design, build and operation of increasingly
complex data center estates, and by transitioning from traditional
Data Center Infrastructure Management (DCIM) to UIIM – ie
reactive to proactive management of operations, data center
owners will be better equipped for a highly competitive and
regulated industry.

CONTENTS

1 Introduction 4

Chapter 1:
2 Leveraging AI and Machine Learning for Predictive Maintenance and 6
Operations Optimization

Chapter 2:
3 Implementing Universal Intelligent Infrastructure Management (UIIM) 21
for Enhanced Operational Efficiency

Chapter 3:
4 Sustainability Reporting and EU Legislation – Navigating Compliance 27
and Operational Challenges

Appendix 1 About the Authors and Contributors 36

Appendix 2 Companies that Participated 38

3
 AI, UIIM, AND ALL THINGS SUSTAINABILITY

INTRODUCTION This white paper presents the findings of the Thirst for Innovation (TFI)
executive roundtable held at Savoy IET on September 26, 2024. Bringing
together a diverse group of over 30 delegates including owner-operators,
supply chain members, and other professionals in the data center
industry, the participants, detailed in Appendix 2, explored challenges,
opportunities, and actionable strategies across the following roundtable
discussions:

There were three round table discussions:

Roundtable Discussion 1:
Leveraging AI and Machine Learning for Predictive Maintenance
and Operations Optimization

Roundtable Discussion 2:
Implementing Universal Intelligent Infrastructure Management (UIIM)
for Enhanced Operational Efficiency

Roundtable Discussion 3:
Sustainability Reporting to Meet EU Legislation

From these discussions, the authors of this white paper have identified
several areas for continuous improvement, offering transparent and
actionable insights to guide progress. In doing so, the authors underscore
the importance of an all-encompassing approach to sustainability,
operational efficiency, and technological integration—critical elements for
effective data collection and management within data center facilities.

Opening with a discussion about the potential of AI and machine learning,

Chapter 1 demonstrates how these technologies are revolutionizing
predictive maintenance and operational optimization. As discussed, though
barriers to adoption related to trust, data standardization, legacy system
integration, and staff shortages remain, the fact that these technologies
enable greater efficiency, reliability, sustainability, and resilience will likely
make them indispensable from an Environmental and Social Governance
(ESG) and fiscal point of view, especially in this evolving industry landscape.

4
AI, UIIM, AND ALL THINGS SUSTAINABILITY

Expanding on this discussion, Chapter 2 provides a more precise

discussion of the Universal Intelligent Infrastructure Management
(UIIM) framework. Designed to unify IT and facility operations into a
streamlined management platform, this framework provides the AI-driven
insights, predictive analytics, and automation necessary for enhanced
decision-making and resource allocation. In doing so, UIIM harbors a
more holistic approach to infrastructure oversight and effective power
utilization measurement. Though barriers including the need for workforce
education, data standardization, and vendor-agnostic integration, and
cultural silos between IT and Operations teams remain, the benefits
are substantial, especially if implementation is phased in through cross-
functional collaboration and the showcasing of long-term value beyond
mere cost savings.

Following on from that, Chapter 3 examines the critical importance of

sustainability reporting in aligning operations with new transparency
requirements set out by the EED and CSRD, highlighting both the
challenges and opportunities this presents. As discussed, though the
robust requirements may initially seem daunting, especially when
accounting for outdated and old servers, poor metering, and IT/facility
silos, Addressing these issues requires robust governance, better
reporting tools, and cross-team collaboration.

Across all of these discussions, there is a general advocation of a shift from

reactive to proactive management strategies for optimized performance,
reduced environmental impact, and the assurance of long-term resilience
in a rapidly evolving and increasingly demanding industry.

5
 CHAPTER 1

LEVERAGING AI AND MACHINE LEARNING

FOR PREDICTIVE MAINTENANCE AND
OPERATIONS OPTIMIZATION
JEFF SAFOVICH, CTO

6
AI AND ML FOR PREDICTIVE OPERATIONS OPTIMIZATION

INTRODUCTION: The data center industry is undergoing a significant transformation driven

THE AI REVOLUTION IN by advancements in Artificial Intelligence (AI) and Machine Learning
(ML), both of which are integral to a Universal Intelligent Infrastructure
DATA CENTER OPERATIONS
Management paradigm (UIIM) described more in depth in Chapter 2.
Traditionally, data center operations have relied on reactive approaches,
where issues are addressed only after they occur. This method often
leads to unexpected downtime, increased maintenance costs, and
inefficient resource utilization, underscoring the need for a more unified
approach—one that UIIM principles address by aligning AI-driven
capabilities across all infrastructure layers. However, the integration of
AI and ML is shifting the paradigm from a reactive to a proactive and
predictive model, fundamentally enhancing how data centers operate.

AI and ML technologies enable data centers to analyze vast amounts of

data in real-time, identifying patterns and predicting potential issues
before they escalate into critical problems. This predictive capability not
only minimizes downtime but optimizes maintenance schedules, ensuring
that equipment is serviced only when necessary. Additionally, AI-driven
automation streamlines complex processes, reducing the need for manual
intervention and allowing data center staff to focus on more strategic tasks.

Efficiency and sustainability are becoming increasingly important in

the data center industry. AI plays a pivotal role in optimizing energy
consumption, managing cooling systems, and balancing workloads
across servers, promoting a holistic and adaptable approach to resource
management. By intelligently allocating resources based on demand
and operational conditions, AI helps data centers reduce their carbon
footprint and achieve sustainability goals without compromising
performance. This significant shift towards intelligent infrastructure
management promises greater efficiency, reliability, and sustainability.

The AI revolution is also enabling innovation in areas such as edge computing

and dynamic resource allocation. As businesses increasingly rely on edge
deployments to support real-time applications and services, AI’s ability
to maintain reliability and efficiency in these distributed environments
becomes crucial. Furthermore, dynamic load balancing powered by
AI ensures that workloads are evenly distributed across data centers,
preventing overloading and enhancing overall system resilience.

In summary, the integration of AI and ML into data center operations

is not just an incremental improvement but a major shift that promises
greater efficiency, reliability, and sustainability. As the industry continues
to evolve, embracing these technologies will be essential for data centers
to stay competitive and meet the growing demands of the digital age.

7
 AI AND ML FOR PREDICTIVE OPERATIONS OPTIMIZATION

AI AND ML APPLICATIONS AI and ML are revolutionizing various aspects of data center operations,
IN DATA CENTER offering innovative solutions that enhance efficiency, reliability, and
sustainability. These solutions become far more impactful when they
OPERATIONS
operate within a Universal Intelligent Infrastructure Management (UIIM)
framework, which emphasizes seamless data sharing and proactive
orchestration, offering innovative solutions that enhance efficiency,
reliability, and sustainability. This section delves into the key applications
of AI and ML in data centers, exploring how these technologies are
being leveraged to optimize operations and drive industry evolution.

Automation of Application and Hardware Deployment

Deploying applications and hardware in a data center involves coordinating

multiple factors such as spatial constraints, power availability, cooling
requirements, and network configurations. AI streamlines this complex
process by automating decision-making based on real-time data and
predefined parameters. For instance, when deploying a new server, AI can
analyze current resource utilization, predict future demand, and determine
the optimal placement to balance load and minimize energy consumption.
This automated orchestration aligns with the principles of UIIM, streamlining
complex deployment processes across the entire data center ecosystem.

Predictive Maintenance for Critical Equipment

Within a UIIM-aligned environment, maintaining the critical infrastructure

of a data center (e.g., UPS, chillers, cooling systems) becomes a proactive,
data-driven effort—essential for uninterrupted operations and streamlined
resource allocation. Traditional maintenance schedules are often based
on fixed intervals, which may not align with the actual condition of the
equipment. In a paradigm of intelligent infrastructure management,
AI and ML algorithms can analyze historical and real-time data from
sensors to predict potential failures with higher accuracy. For example,
by monitoring vibration patterns and temperature readings, AI can
forecast when a UPS unit is likely to fail, allowing for timely maintenance
that prevents unexpected downtime and reduces repair costs. In edge
computing applications, this predictive maintenance is crucial for ensuring
the reliability of distributed systems without frequent on-site visits.

8
AI AND ML FOR PREDICTIVE OPERATIONS OPTIMIZATION

Integration Challenges and the Role of AI

The disparity in protocols and data formats across equipment vendors

and individual modules poses a significant challenge to achieving the
holistic integration envisioned by UIIM. AI-powered tools are essential
for facilitating cross-system orchestration, enabling integration with
diverse physical equipment and software systems. By leveraging AI for
data unification, effective inter-system communication, and automation of
workflows, data centers can streamline operations and enhance efficiency.

Unified data from all sources and systems is a prerequisite for implementing
intelligence-driven capabilities such as predictions, machine learning,
proactive recommendations, and automation. This integration requires
advanced technologies that can not only read data but also enable
effective cross-system workflows. Within UIIM frameworks, AI acts as an
orchestrating intelligent arbiter—facilitating cross-system ‘communication’
and automating routine tasks—to enhance operational efficiency and
reduce the risk of human error. This universal data model is central to
UIIM’s goal of bringing all infrastructures under a single pane of glass.

Edge Computing and Telecommunications Optimization

Edge computing brings data processing closer to the source, reducing

latency and improving the performance of applications that require
real-time responses. AI enhances edge deployments by optimizing
telecommunications and maintenance processes. By predicting when and
where maintenance is needed, AI reduces the number of on-site visits,
lowering operational costs and minimizing disruptions. Additionally, AI can
manage network traffic efficiently, ensuring that data is routed optimally
to maintain high performance and reliability. This optimization is vital
for applications such as autonomous vehicles, smart cities, and Internet
of Things (IoT) devices, where consistent performance is critical.

Optimizing Maintenance Schedules: Filter Replacement Example

Maintenance tasks like filter replacements are essential for maintaining the
efficiency of cooling systems in data centers. However, performing these
tasks too frequently can lead to unnecessary costs and resource wastage,
while delaying them can compromise system performance. AI can analyze
operational data, such as energy usage, airflow patterns, and filter condition
indicators, to determine the optimal intervals for maintenance. For instance,
by monitoring the pressure differential across a filter, AI can predict the
exact moment when a filter needs replacement, ensuring that maintenance
is performed only when necessary. This approach improves efficiency,
extends the lifespan of equipment, and reduces overall maintenance costs.

9
 AI AND ML FOR PREDICTIVE OPERATIONS OPTIMIZATION

Workload Optimization and Dynamic Load Balancing

Efficiently managing workloads across servers is crucial for maintaining

performance and preventing resource bottlenecks. AI can play a
significant role in dynamically balancing loads by analyzing real-time
data on server utilization, application demands, and network conditions.
When an imbalance is detected, AI can automatically redistribute
workloads to ensure optimal performance. For example, if one server
is experiencing high CPU usage while others are underutilized, AI can
migrate some of the workloads to the less busy servers. This dynamic
load balancing, a cornerstone of intelligent infrastructure management,
enhances system reliability and ensures efficient resource utilization.
Integrating AI for workload management does present challenges, such
as ensuring seamless migration and maintaining data consistency, but
the benefits in operational efficiency make it a worthwhile investment.

Energy Efficiency and Sustainability through AI

Energy consumption is a major concern for data centers, both in terms

of operational costs and environmental impact. AI can contribute
to energy efficiency by optimizing cooling systems, power usage,
and resource allocation in real-time. For example, AI algorithms can
adjust cooling parameters based on current server loads and ambient
temperatures, ensuring that cooling resources are used only where and
when needed. Additionally, AI can manage power distribution to avoid
overloading circuits and reduce energy waste. By continuously analyzing
data from various sensors, AI helps data centers minimize their energy
consumption, lower operational costs, and achieve sustainability targets.
This optimization not only benefits the environment but also enhances
the overall efficiency and competitiveness of data center operations.

Dynamic Resource Allocation for Scalability

As businesses grow and their data processing needs increase, data centers
must scale their resources efficiently to handle larger workloads without a
proportional rise in energy consumption or operational complexity. AI can
facilitate this scalability by dynamically allocating resources based on real-
time demand. For instance, during peak usage periods, AI can automatically
provision additional computing power or storage capacity to meet the
increased demand. Conversely, during periods of low activity, AI can scale
back resources to conserve energy. This dynamic allocation ensures that
data centers can handle fluctuating workloads seamlessly, maintaining high
performance while optimizing resource utilization. Additionally, AI can predict
future resource needs based on usage trends, enabling proactive scaling
that supports business growth without incurring unnecessary costs.

10
AI AND ML FOR PREDICTIVE OPERATIONS OPTIMIZATION

CHALLENGES AND While the integration of AI and ML in data center operations offers
BARRIERS TO AI ADOPTION numerous benefits, several challenges and barriers must be addressed
to ensure successful adoption—challenges that UIIM methodologies
anticipate by advocating universal data standards, comprehensive
security protocols, and integrated AI governance. This section
explores the key obstacles organizations face when implementing AI
solutions and provides insights into overcoming these hurdles

Building Trust and Ensuring Reliability

One of the primary challenges in adopting AI systems is building trust

among stakeholders. Data center operators must have the confidence
that AI-driven decisions are accurate and reliable. To establish this trust,
it is crucial to ensure transparency in how AI models make decisions.
This involves providing clear explanations of AI processes and outcomes,
allowing operators to understand and verify the rationale behind AI
recommendations. Additionally, demonstrating consistent performance
through rigorous testing and validation can help prove the reliability of
AI systems. By prioritizing transparency and reliability, organizations will
be able to trust and encourage broader adoption of AI technologies.

Navigating Security Concerns in AI Deployments

AI deployments introduce unique security challenges that data centers

must address. AI systems often require access to sensitive data, which
can become a target for cyberattacks. Ensuring robust security measures
tailored to AI environments is essential to protect against vulnerabilities.
This includes implementing advanced encryption techniques, secure data
storage solutions, and continuous monitoring for suspicious activities.
Moreover, AI systems themselves can be vulnerable to attacks such as
adversarial inputs, where malicious actors manipulate data to deceive the
AI. Developing comprehensive security protocols that encompass both the
data and the AI models is vital for safeguarding data center operations.

Data Integrity, Standardization, and Silos

AI’s effectiveness relies heavily on the quality and consistency of the

data it processes. However, data integrity issues, such as incomplete or
inaccurate data, can significantly hinder AI performance. Additionally,
data often resides in silos across different departments or systems,
making it challenging to create a unified dataset for AI analysis. To
overcome these challenges, organizations must invest in data quality
initiatives, ensuring that data is accurate, complete, and in real time.
Establishing standardized data formats and promoting data sharing
across the organization can help break down silos and create a cohesive
data environment. Collaboration across departments is essential to
maintain high data standards and support effective AI implementation.

11
 AI AND ML FOR PREDICTIVE OPERATIONS OPTIMIZATION

Integrating AI with Legacy Infrastructure

Many data centers operate with legacy systems that are not designed
to support modern AI technologies. Integrating AI with these older
infrastructures can be complex and resource-intensive, highlighting
the need for a flexible and adaptable management framework that can
bridge the gap between old and new. One strategy to address this is
adopting a phased integration approach, where AI solutions are gradually
introduced alongside existing systems. This allows organizations to
test and refine AI applications without disrupting current operations.
Additionally, developing AI solutions that are compatible with legacy
infrastructure can facilitate smoother integration. Investing in middleware
or APIs that bridge the gap between old and new technologies can
also help streamline the integration process, enabling data centers to
leverage AI benefits without overhauling their entire infrastructure.

Developing a Regulatory Framework for AI in Data Centers

The rapid advancement of AI has outpaced the development of

regulatory frameworks tailored to its unique challenges. Unlike other
regulated industries, the data center sector lacks specific guidelines
governing AI usage, which can lead to inconsistencies and potential
risks. Establishing a comprehensive regulatory framework for AI in data
centers is essential to ensure responsible adoption. This framework
should address issues such as data privacy, ethical AI use, transparency
requirements, and accountability measures. By drawing on best practices
from other regulated industries and collaborating with stakeholders,
organizations can help shape regulations that promote safe and effective
AI deployment, creating an environment of trust and innovation.

Maintaining Human Oversight in an AI-Driven World

Despite the increasing capabilities of AI, maintaining human oversight

remains crucial in data center operations. AI should be viewed as a
tool that augments human decision-making rather than replacing it
entirely. Human operators bring contextual understanding, ethical
considerations, and adaptability that AI currently lacks. To balance
AI automation with human control, organizations should implement
governance structures that define the roles and responsibilities of both AI
systems and human staff. Regular audits of AI performance and decision-
making processes can ensure that AI remains aligned with organizational
goals and ethical standards. By preserving human oversight, data
centers can leverage AI’s strengths while mitigating its limitations.

12
AI AND ML FOR PREDICTIVE OPERATIONS OPTIMIZATION

Addressing the Skill Gap and Training Needs

The successful implementation of AI in data centers requires a workforce

with the necessary skills to manage and interact with AI systems. However,
there is often a significant skill gap, with many data center staff lacking
expertise or knowledge of AI and ML technologies, physical assets and
infrastructure. To address this, organizations must invest in comprehensive
training and skill development programs. These programs can include formal
education, hands-on workshops, and continuous learning opportunities to
help employees understand AI concepts, operate AI tools, and interpret
AI-driven insights. Creating a culture of continuous improvement and
encouraging cross-functional collaboration will help to bridge the skill gap.

THE FUTURE OF AI IN DATA Looking ahead, AI is set to play an even more transformative role in data
CENTERS: A VISION OF center operations. This section envisions the future landscape where AI
technologies enhance decision-making, automate processes, and drive
ENHANCED OPERATIONS
innovation, ultimately leading to more efficient, resilient, and sustainable
data centers.

AI-Enhanced Decision-Making for Complex Operations

AI will increasingly provide data-driven insights and recommendations,

empowering data center managers to make more informed decisions.
By analyzing vast amounts of operational data within a unified UIIM
ecosystem, AI can identify trends, predict potential issues, and suggest
optimal strategies for resource allocation and risk management—
bolstering not just operational data but the entire infrastructure’s
cohesiveness. This enhanced decision-making capability enables
managers to respond swiftly to changing conditions, improving overall
operational efficiency and reducing the likelihood of costly mistakes.

Increased Automation and Operational Efficiency

Automation will continue to expand, with AI taking on more routine tasks

and processes within data centers. This increased automation frees human
operators to focus on strategic initiatives and innovation, rather than being
bogged down by repetitive tasks. For example, AI can handle tasks like system
monitoring, routine maintenance, and even complex processes like workload
balancing. As AI manages these aspects with minimal human intervention, data
centers can achieve higher levels of operational efficiency and productivity.

13
 AI AND ML FOR PREDICTIVE OPERATIONS OPTIMIZATION

Transitioning to Proactive and Predictive Management

The shift from reactive problem-solving to proactive management will be

a key development in the future of data centers. AI’s ability to predict and
prevent issues before they impact operations allows data centers to maintain
higher levels of reliability and uptime. Predictive management strategies
enable organizations to anticipate and mitigate potential disruptions,
ensuring smooth and uninterrupted service delivery.

AI-Driven Scalability for Future Growth

As data centers need to scale to meet growing demands, AI will play a crucial
role in managing this growth efficiently. By dynamically allocating resources
based on real-time demand, AI ensures that data centers can handle
increasing workloads without a corresponding rise in energy consumption or
operational complexity. This intelligent scaling capability allows data centers
to expand seamlessly, supporting business growth while maintaining optimal
performance and cost-efficiency.

Enhanced Cybersecurity with AI-Powered Threat Detection

Cybersecurity will be significantly bolstered by AI technologies. AI can analyze

network traffic patterns, detect anomalies, and identify potential threats in
real-time, providing a proactive defense against cyberattacks. Additionally, AI
can automate incident response processes, enabling faster and more effective
mitigation of security breaches. This enhanced threat detection and response
capability ensures that data centers remain secure in the face of evolving
cyber threats.

Achieving Sustainability Goals through AI Optimization

AI will be instrumental in helping data centers achieve their sustainability

goals. By optimizing energy consumption, reducing waste, and integrating
renewable energy sources, AI contributes to making data center operations
more environmentally friendly. For instance, AI can adjust cooling systems
based on real-time data to minimize energy usage or manage power
distribution to maximize the use of renewable energy. These optimizations
not only reduce the environmental impact but also lead to cost savings and
improved operational efficiency.

14
AI AND ML FOR PREDICTIVE OPERATIONS OPTIMIZATION

The Path to Autonomous Data Centers

The long-term vision for data centers involves achieving near-autonomous

operations, where AI manages almost all aspects of the center’s functionality.
In such autonomous data centers, AI systems would handle everything
from workload distribution and energy management to security and
maintenance. This level of automation would result in highly efficient, resilient,
and self-managing facilities that require minimal human intervention,
allowing organizations to focus on innovation and strategic growth.

Fostering Collaboration and Standardization for AI Advancement

For AI to reach its full potential in data center operations, collaboration and
standardization across the industry are essential. Data center operators,
technology vendors, and regulatory bodies must work together to establish
standards, best practices, and guidelines for AI implementation. This
collaborative approach ensures interoperability between different AI systems
and promotes the responsible and effective use of AI technologies. By
creating a unified framework, the industry can accelerate AI advancements
and ensure consistent quality and reliability across data centers.

Unlocking Innovation and New Capabilities with AI

AI will drive significant innovation in various aspects of data center operations,

opening up new capabilities and business opportunities. For example, AI can
enhance disaster recovery strategies by predicting potential failures and
automating recovery processes. In data analytics, AI can provide deeper
insights into operational performance, enabling continuous improvement.
Additionally, AI can support the development of new business models, such
as offering AI-as-a-service to clients or creating more flexible and scalable
service offerings. These innovations not only improve data center operations
but also provide a competitive edge in the rapidly evolving digital landscape.

15
 AI AND ML FOR PREDICTIVE OPERATIONS OPTIMIZATION

ROUNDTABLE INSIGHTS During the roundtable focusing on AI and ML for predictive

AND RECOMMENDATIONS maintenance and operations optimization, participants offered a
range of practical observations, lessons learned, and forward-looking
suggestions. The discussion underscored both the immediate benefits
of AI as well as the broader organizational and cultural shifts needed
to support AI-driven data center operations. Below is a distilled
overview of key insights and actionable recommendations:

Embrace Incremental Adoption and Phased Integration

Insights:
P
 articipants noted that AI adoption does not require a total overhaul of
existing infrastructure or operations. Instead, incremental steps—such
as using AI for predictive maintenance or energy optimization—fit neatly
into a UIIM-based phased strategy, where universal data models and
integrative workflows enable quick demonstrations of tangible benefits.

E
 arly “small wins” help build trust in AI, especially in cautious
environments where critical operations are at stake.

Recommendations:
 tart with specific use cases (e.g., filter replacement scheduling, UPS
S
performance monitoring) before scaling to broader operations.
 ilot in low-risk areas to validate AI’s value and
P
build organizational momentum.
 xpand as confidence grows, positioning AI to handle more
E
critical and complex tasks once proven at smaller scales.

Build and Maintain Trust in AI

Insights:
T
 he group underscored trust as both a technical and cultural challenge.
Operators need confidence that AI systems will make reliable, explainable
decisions—especially regarding critical infrastructure like power and cooling.
T
 rust grows from transparency: data center teams must
understand how AI makes its recommendations and be able
to validate outcomes against real-world events.

Recommendations:
 dopt “explainable AI” approaches so operators can
A
audit and understand AI-driven decisions.
 ombine AI-driven insights with human oversight, ensuring
C
that critical decisions (like scheduling maintenance or rerouting
workloads) are confirmed by experienced staff.
 hare success stories and lessons learned across teams and
S
departments to demystify AI’s “black box” reputation.

16
AI AND ML FOR PREDICTIVE OPERATIONS OPTIMIZATION

Integrate AI Securely and Address Data Quality

Insights:
A
 I solutions are only as good as the data they consume; poor data
integrity or fragmented data silos lead to flawed insights.
P
 articipants warned that increased AI adoption can introduce new attack
vectors, as AI systems often require broad access to operational data.
Recommendations:
 stablish robust data governance to ensure data is accurate,
E
consistent, and accessible in secure, standardized formats.
I mplement layered security for AI deployments (e.g., strong access
controls, continuous monitoring, and encryption for AI-related data).
 ollaborate with vendors and peers to work toward
C
standard data models and protocols, reducing the friction
of integrating AI across multiple platforms.

Use AI to Shift from Reactive to Proactive Operations

Insights:
O
 ne of the most tangible benefits of AI is the ability to move
away from “break-and-fix” or purely scheduled maintenance
models to predictive and condition-based strategies.
M
 any participants viewed predictive maintenance and real-time workload
balancing as “low-hanging fruit” for demonstrating AI’s effectiveness.
Recommendations:
D
 eploy AI for predictive analytics around key assets (UPS, chillers, cooling
infrastructure), leveraging historical performance data to anticipate failures.
O
 ptimize maintenance intervals with AI that tracks
operational data (e.g., energy consumption, temperature
fluctuations) to reduce unnecessary service calls.
E
 xplore workload optimization for dynamically distributing
tasks across facilities or even shifting workloads
during peak demand to prevent bottlenecks.

17
 AI AND ML FOR PREDICTIVE OPERATIONS OPTIMIZATION

Focus on Energy Optimization and Sustainability

Insights:
P
 articipants consistently highlighted energy optimization—particularly
around cooling—as a critical area where AI can make a measurable
impact under the UIIM. By unifying real-time performance metrics,
data centers can leverage predictive analytics to reduce consumption,
manage renewable resources, and comply with emerging regulations.
R
 egulatory pressures and corporate sustainability goals
are growing. AI-driven insights can reduce consumption,
manage renewable resources, and cut costs.
Recommendations:
I ntegrate AI into cooling management, using real-time data
to adjust fan speeds, temperature setpoints, or cooling
circuits based on workload and ambient conditions.
 tilize AI-powered forecasting to anticipate workload spikes and
U
preemptively adjust energy allocations (e.g., ramping up cooling or
shifting resources to locations with green energy availability).
 rack and report energy metrics for both cost savings and compliance
T
with emerging sustainability legislation. AI-based reporting can
offer more granular insights than traditional methods.

Address Skills Gaps and Human-AI Collaboration

Insights:
D
 ata center staff often need additional training to interpret AI-
driven recommendations and to maintain AI systems responsibly.
R
 ather than replacing staff, AI augments human capabilities, freeing
operators to focus on strategic tasks and higher-level problem-solving.
Recommendations:
I nvest in training and professional development, ensuring teams
can utilize AI outputs effectively and maintain AI tools.
 romote collaboration between IT, operations, and data science
P
teams for continuous learning and experimentation.
 edefine roles and responsibilities to reflect a more
R
AI-augmented environment, ensuring that human
expertise remains integral to final decisions.

18
AI AND ML FOR PREDICTIVE OPERATIONS OPTIMIZATION

Plan for Future Scale and Regulatory Compliance

Insights:
 I can help data centers remain agile as demand grows, but systems
A
must be designed with scalability in mind from the outset.
 egulations around data usage, security, and sustainability
R
are tightening. AI can help with compliance reporting,
but only if governance frameworks are in place.
Recommendations:
 dopt scalable AI architectures that can handle growing
A
workloads and integrate with legacy equipment over time.
 onitor evolving regulations to align AI deployments with emerging
M
data governance, cybersecurity, and sustainability requirements.
 roactively engage with industry bodies and regulators,
P
sharing best practices and helping shape sensible
frameworks that support responsible AI adoption.

Look Ahead to Full or Partial Autonomy

Insights:
 hile fully autonomous data centers remain a future aspiration, several
W
participants believed that a hybrid model—where AI manages day-to-day
operations with human oversight—will become increasingly common.
 ver time, as AI proves its reliability, data centers may
O
trust AI to self-manage tasks such as workload distribution,
energy throttling, or routine maintenance scheduling.
Recommendations:
 everage AI to automate routine tasks, then gradually
L
expand autonomy once trust is established.
 et clear escalation procedures detailing when and how human
S
intervention occurs if AI systems flag anomalies or critical events.
 xperiment with “lights-out” scenarios (short periods or certain functional
E
areas) to test the feasibility of higher levels of AI-driven autonomy.

19
 AI AND ML FOR PREDICTIVE OPERATIONS OPTIMIZATION

CONCLUSIONS: The integration of AI and ML into data center operations marks a profound
THE TRANSFORMATIVE shift from reactive to predictive and proactive management, promising
enhanced efficiency, reliability, and sustainability. Insights from the Thirst
POTENTIAL OF AI IN DATA
For Innovation roundtable highlight the transformative potential of AI
CENTER OPERATIONS across predictive maintenance, energy optimization, dynamic resource
allocation, and enhanced decision-making. While challenges such as data
integrity, security, and skills gaps remain, these can be addressed through
phased integration, transparent processes, and industry collaboration.

As the industry continues to evolve, the adoption of AI will not only

redefine operational paradigms but also enable data centers to meet
growing demands, comply with regulatory standards, and achieve
ambitious sustainability goals. This sets the stage for Chapter 2’s deeper
exploration of UIIM as the next evolution of data center management.

By embracing AI strategically and developing trust among stakeholders,

data centers can pave the way for innovation, resilience, and future-ready
operations. The journey towards more autonomous data centers is well
underway, laying a foundation for a smarter, more sustainable digital future.

20
CHAPTER 2

ADOPTION OF UNIVERSAL INTELLIGENT

INFRASTRUCTURE MANAGEMENT (UIIM)
FOR ENHANCED OPERATIONAL EFFICIENCY
ANTONIO SUAREZ, GLOBAL PRODUCT MANAGER

21
 WHAT IS UIIM?

WHAT IS UNIVERSAL The Universal Intelligent Infrastructure Management (UIIM) framework,

INTELLIGENT introduced by RiT Tech in 2020, represents a significant shift in data center
infrastructure management (DCIM), offering capabilities far beyond those
INFRASTRUCTURE
of traditional DCIM. Initially introduced by Gartner in 2009, DCIM aimed to
MANAGEMENT? bridge IT and Mechanical & Electrical (M&E) infrastructure management.
However, it failed to fully integrate the complex needs of modern data
centers, particularly as data center environments evolved into more
diverse forms such as colocation, enterprise, and modular setups. While
DCIM systems provided limited monitoring capabilities, they were never
designed to manage today’s diverse data centers, encompassing everything
from enterprise facilities to colocation and modular configurations.

In response, UIIM has emerged as a comprehensive, adaptable approach

to infrastructure management, leveraging real-time data insights,
predictive analytics, and AI-driven automation to manage the entire
operational ecosystems. Unlike traditional DCIM, which focuses mainly
on hardware monitoring and specific infrastructure components,
UIIM’s holistic approach supports seamless integration across systems
and platforms, aligning with the industry’s need for scalable, flexible
management tools that go beyond simple monitoring to actively optimize
performance. This chapter examines the adoption challenges for UIIM and
its major potential as a solution for modern data center environments.

CORE FEATURES OF UIIM UIIM distinguishes itself from DCIM with core features enabling it to
manage complex, multifaceted data center operations effectively. First and
foremost, UIIM proposes a universal platform that works across various data
center configurations, including colocation, modular, Edge, and enterprise
setups. By creating a single interface, UIIM provides a unified and real time
view of all data center components, both IT and M&E, allowing operators to
monitor and control the entire infrastructure from one centralized platform.
This universal compatibility greatly simplifies operations, as data centers
no longer need to manage multiple fragmented systems, enabling a “single
pane of glass” experience that enhances efficiency and responsiveness.

Another defining feature within the UIIM paradigm is its AI-driven

intelligence, which enables predictive analytics, accurate resource planning,
and adherence to regulatory compliance. Where DCIM tools relied on
static, predefined thresholds, UIIM’s AI processes adapts in real-time to
changing conditions, enabling data centers to predict potential failures
before they escalate and adjust resource allocation dynamically.

UIIM’s ability to scale with a data center’s growth is also critical. Unlike
DCIM, which often fell short in environments requiring flexible scalability,
UIIM’s adaptable framework enables it to manage resources and workloads
dynamically, responding to fluctuations in demand and predicting capacity
needs based on historical trends. This real-time adaptability reduces
the need for extensive manual intervention, avoiding the bottlenecks

22
UIIM IMPLEMENTATION - ENHANCED OPERATIONAL EFFICIENCY

traditionally accompanying expansion efforts. As the industry shifts towards

cloud-based and hybrid infrastructures, UIIM’s scalability ensures that data
centers can meet rising demand without compromising operational stability.

The benefits of UIIM are far-reaching, impacting everything from

operational efficiency to sustainability and regulatory compliance. One of
the most significant advantages of UIIM is its function as a single source
of truth. Unlike traditional systems that rely on disparate data sources,
UIIM consolidates information across IT and facility infrastructure,
creating a “golden source” of real-time data that ensures accurate, reliable
insights. This unified data source supports enhanced decision-making, as
operators can trust that their decisions are based on consistent, up-to-
date information, whether managing workloads, monitoring asset health, or
responding to facility alarms.

In addition to centralized data management, UIIM enables process

optimization and predictive maintenance. By analyzing historical and real-
time data across systems, UIIM identifies trends that indicate potential
issues, allowing data centers to proactively address them before they
escalate. For example, suppose UIIM detects patterns that suggest a
cooling system will require maintenance within the next month. In that case,
operators can schedule repairs during planned downtime, reducing the
risk of unexpected failures. This predictive approach improves operational
resilience and extends critical equipment’s life cycle, reducing repair costs
and downtime. Furthermore, UIIM’s automation capabilities streamline
routine tasks, freeing up resources and allowing staff to focus on high-value
activities that drive innovation.

Regulatory compliance is another area in which UIIM provides substantial

value. With increasing scrutiny on energy consumption, emissions, and
environmental impact, data centers face pressure to meet industry and
regional regulations. UIIM simplifies compliance by automating reporting
processes and ensuring that data centers operate within the parameters
set by regulatory bodies. By continuously monitoring metrics such as
energy use, cooling efficiency, and emissions, UIIM ensures that facilities
remain compliant while identifying areas where sustainability efforts can be
enhanced. This proactive compliance management helps data centers avoid
penalties, improve their environmental footprint, and build a competitive
reputation as responsible, sustainable operations.

Finally, UIIM breaks down the silos that often hinder collaboration in
data centers. IT and facility management have traditionally operated
independently, with limited data sharing. UIIM bridges this gap by
consolidating all data into a single interface, enabling a holistic view of IT
and facility operations. This integrated approach fosters cross-functional
collaboration, allowing IT and facility teams to work together to resolve
issues that affect both systems. By creating a collaborative environment,
UIIM helps data centers reduce inefficiencies, enhance resource allocation,
and achieve better alignment between IT and facility objectives.

23
 UIIM IMPLEMENTATION - ENHANCED OPERATIONAL EFFICIENCY

WHAT ARE THE

ADOPTION BARRIERS TO ADOPTION
CHALLENGES FOR 1. Lack of understanding the benefits of UIIM could slow down
UIIM? adoption. Use cases will be required for better education.

2. IT and facilities teams must collaborate to enable a unified

data set.

3. A standardized approach to collection and management

of coherent data will enable industry-wide deployment.

4. Vendor-specific systems require integration with UIIM,

requiring hardware vendors to adopt.

5. Adoption of UIIM faces barriers like upfront costs, ROI

concerns, cultural resistance, and operational challenges
despite clear long-term benefits.

The roundtable discussion identified that inadequate education about

the benefits of Universal Intelligent Infrastructure Management (UIIM)
compared to Data Center Infrastructure Management (DCIM) is a
significant barrier in the data center industry. UIIM is still in its early
stages of adoption and its success hinges on the industry’s willingness to
embrace substantial shifts in collaboration and support its path toward
a more regulated framework. This specific discussion was designed to
build an understanding among industry leaders to prepare a strategy
for building industry-wide awareness for this innovative framework.

Overcoming Silos: IT and Facilities Collaboration

One major obstacle is the cultural divide between IT and Operational

Technology (OT) teams. Historically, these teams have operated in silos,
with limited understanding of each other’s roles and priorities. UIIM
necessitates a more integrated approach, fostering collaboration to
bridge the gap between IT and facilities teams. This synergy is critical
to achieving accurate, actionable insights that enhance operational
efficiency. Investing in training programs designed to facilitate
collaboration is one way to address this challenge. By empowering
teams with the skills to work cohesively, data centers can maximize
the benefits of UIIM and build more agile, resilient operations.

Standardization: The Key to Consistency

Another challenge is the lack of standardized data and terminology

across systems. Data centers often rely on tools from multiple vendors,
leading to inconsistent labeling and measurement conventions
for identical components. UIIM tackles this issue by introducing
standardized data conventions and naming protocols, enabling
seamless communication and a unified view of infrastructure. However,
implementing these standards requires careful coordination across
internal teams and external vendors. Standardization efforts, including
consistent data labeling and terminology, create a cohesive system

24
UIIM IMPLEMENTATION - ENHANCED OPERATIONAL EFFICIENCY

that simplifies decision-making and enhances collaboration.

Interoperability: Bridging Vendor Ecosystems

As a universal interface, UIIM addresses interoperability challenges

by offering a vendor-agnostic view of all infrastructure components.
However, fully integrating diverse systems requires overcoming
inconsistencies in data formats, protocols, and compatibility. Achieving
seamless functionality demands proactive planning and robust
collaboration between internal teams and external partners. Ensuring that
UIIM interfaces smoothly with all vendor systems is critical to its success,
as it provides a foundation for effective, standardized operations.

Upfront Costs: Changing Mindsets from CapEx to OpEx priorities

While the long term benefits of UIIM are clear, adoption can be
hindered by the costs associated with deployment and integration.
Some CFOs and CEOs concerned with an immediate ROI, making
adoption difficult. It will be critical to show use cases where UIIM
improves maintenance costs, data center resilience, and how it
impacts on operational efficiency. UIIM has longer term benefits such
as informing design through techniques such as digital twin, process
optimization to inform protocols, and regulatory compliance.

UIIM requires data centers to look beyond short-term financial

metrics and consider reducing Total Cost of Ownership (TCO)
and moving from capital expenditure (CapEX) intensive models
to more flexible operational expenditure (OpEX) models.

KEY REGULATION AND As the data center industry evolves, several trends will support the
adoption and implementation of UIIM. As data centers face growing
INDUSTRY TRENDS
pressure to improve their environmental impact, the future of UIIMs
TOWARDS REMOTE success lies in enabling sustainable, intelligent operations. UIIM’s capacity
MANAGEMENT WILL for real-time data analysis and reporting allows data centers to monitor
BOOST ADOPTION OF UIIM critical metrics related to sustainability, such as energy use, emissions,
and cooling efficiency. By automating these reporting processes, UIIM
simplifies compliance with environmental regulations, allowing data centers
to align with evolving standards without adding administrative burdens.
COLO customers are adopting tools like Universal Intelligent
Infrastructure Management (UIIM) to remote manage their hosted
IT infrastructure. This includes monitoring and managing their own
servers, storage, networking equipment and connectivity, and in-
rack power distribution systems. The industry is moving towards
extending the benefits of UIIM to cover COLO’s Mechanical and
Electrical infrastructure serving hosted IT infrastructure including
cooling, UPS, PDUs and environmental monitoring such as temperature
and humidity, which in practice this means is that through UIIM, the
hosted client manages the entire infrastructure, as if it were hosted in
their own Datacenter. UIIM provides accurate, real-time insights into
critical resources such as power usage, people and other risks such as
management of risks provided by advanced infrastructure and assets.

25
 UIIM IMPLEMENTATION - ENHANCED OPERATIONAL EFFICIENCY

ROUNDTABLE UIIM’s ability to enable cross-functional collaboration is a key driver:

RECOMMENDATIONS To bridge the gap between IT and Operational Technology (OT) teams will
result in better operational management, enhanced customer service and
improved adherence to compliance and protocols.

Invest in UIIM awareness programmes and analyst adoption:

Implement comprehensive awareness programs to enhance understanding
of UIIM’s features, focusing on its integration and predictive capabilities.
Analysts, operators and the entire ecosystem will need to participate for
its adoption to be successful. Training progams can be one major step
forward.

Standardize terminology and data collection protocols:

Establish consistent naming conventions and data protocols across
vendors and systems to streamline operations and facilitate seamless
UIIM integration.

Plan for interoperability:

The industry to prioritize vendor-agnostic solutions and ensure meticulous
planning during UIIM implementation to address system compatibility
challenges and avoid roadblocks.

Highlight long-term value:

Communicate the broader benefits of UlIM-such as sustainability,

resilience, and reduced Total Cost of Ownership (TCO)- to align
stakeholders on strategic goals beyond immediate cost savings.

Adopt a phased implementation strategy:

Start with the basic pillars of UIIM, including integrations with other
systems. Having one centralized management system will add value and as
the data improves, UIIM will grow quickly to support advanced automation
and predictive analysis, such as predictive maintenance.

CONCLUSION UIIM has the potential to redefine data center infrastructure management,
providing an integrated, intelligent solution that meets the industry’s
growing demands for efficiency, scalability, and sustainability. By
consolidating systems, bridging IT and facility silos, and incorporating
real-time insights, UIIM supports data centers in achieving optimized,
resilient operations. However, the success of UIIM hinges on addressing
key challenges around education, standardization, and collaboration. Data
centers that embrace UIIM today position themselves at the forefront of a
rapidly evolving industry, prepared to meet the demands of a digital-first,
sustainability-focused future.

26
CHAPTER 3

SUSTAINABILITY REPORTING AND EU

LEGISLATION – NAVIGATING COMPLIANCE
AND OPERATIONAL CHALLENGES
MARK ACTON, CONSULTANT

OVERVIEW
CHALLENGES
OUTCOMES

27
 SUSTAINABILITY REPORTING AND EU COMPLIANCE CHALLENGES

SUSTAINABILITY As the European Union (EU) intensifies its focus on sustainability, data centers
REPORTING IN are navigating new legislative requirements that mandate comprehensive
reporting on energy efficiency and broader ESG obligations. Through
DATA CENTERS
frameworks like the Energy Efficiency Directive (EED), Corporate Sustainability
Reporting Directive (CSRD), and the European Sustainability Reporting
Standards (ESRS), the EU aims to build a detailed understanding of the
environmental footprint of high-energy-consuming data centers, including
energy and water usage. These initiatives present significant operational
challenges but also hold the potential for driving essential industry shifts
toward renewable energy and responsible resource management.

The complexities of achieving compliance extend beyond reporting alone.

Each EU member state can set its own unique reporting requirements,
creating a fragmented regulatory landscape that complicates things for
cross-border operators. Additionally, data centers operating under colocation
models face further obstacles in accessing reliable data from client-owned
assets. Without complete visibility into tenant energy consumption, operators
cannot provide complete and accurate reports, limiting the impact of
sustainability measures and potentially leading to compliance risks.

Aside from regulatory issues, data center expansion faces increasing

resistance from local communities concerned about ecological risks,
including energy usage, emissions, and water consumption. Without
accurate reporting, it is impossible to prove efficiency or positive impacts,
such as supporting the local and national grids. Important initiatives such
as heat reuse partnerships and the expansive use of renewable energy
integration will also go undocumented.

In this study, the Thirst for Innovation roundtable highlighted that

reporting across data center facilities offers additional benefits beyond
regulatory compliance. Although governments like the UK have designated
data centers as Critical National Infrastructure, public resistance to their
expansion persists. This opposition is partly due to a lack of understanding
of the environmental impacts associated with data consumption, which
remains largely invisible to end-users. Addressing this awareness gap is
essential to managing consumer demand for sustainable digital services.
Some stakeholders suggest new approaches, such as implementing a digital
service usage tax similar to carbon taxes on vehicles or introducing energy
labeling for digital activities to display their environmental impact. In-depth
reporting could raise awareness about the impact of data consumption,
encouraging consumers to make greener choices and creating market-
driven incentives for data centers to adopt and report on sustainable
practices. In this context, accurate and transparent reporting is more than a
regulatory hurdle; it can be a fundamental step towards earning public trust
and driving meaningful change across the industry.

This chapter explores data center providers’ sustainability reporting

challenges, focusing on the impact of EU regulations, logistical obstacles,
and potential compliance strategies. Through standardized reporting and
advanced resource monitoring, data centers can better showcase their
commitment to sustainability. These efforts are crucial for meeting EU
targets and driving the industry toward environmentally responsible growth.

28
SUSTAINABILITY REPORTING AND EU COMPLIANCE CHALLENGES

MEETING REGULATORY In a stakeholder call on September 6, 2023, the EU confirmed it

DEADLINES – ENERGY would not extend the Energy Efficiency Directive (EED) reporting
deadline beyond September 15 despite keeping the reporting
EFFICIENCY DIRECTIVE
portal open for a limited grace period. This decision underscores
(EED) - THE CHALLENGES the EU’s urgency for timely compliance. However, data centers
face significant challenges, with the biggest issue surrounding
lack of clarity from the EU on reporting requirements.

A) Lack of Standardization Across EU Countries

One key obstacle to meeting EED standards is the lack of uniformity

in reporting requirements across EU member states. Each country’s
distinct approach to reporting can cause confusion and inconsistencies,
particularly for data centers operating in multiple jurisdictions.
This divergence impedes efforts toward EU-wide standardization
and can hinder achieving ambitious climate goals. Cross-border
operators risk delays or setbacks in their compliance efforts,
which may widen the gap in regional sustainability progress.

Achieving consistent EU-wide sustainability outcomes requires both

standardization and incentivization. Individual countries have adopted
unique compliance protocols, offering varied incentives and penalties.
For example, France incentivizes compliance with discounts on electricity
bills for data reporters, while the Netherlands enforces strict deadlines
without providing benefits. Germany’s stringent regulations are often
attributed to its commitment to environmental leadership, rooted
in ambitious climate policies and a robust public expectation for
corporate accountability. While this strict approach aims to drive rapid
environmental gains, some view it as a hurdle to data center sector
growth, as it increases operational complexity and compliance costs.

The regulatory challenge data centers face resembles the experience

of the glassblowing industry, where regulation brought clarity and
improved industry practices over time. The lengthy standardization
process ultimately led to better environmental outcomes and
competitive standards, setting a precedent for how cohesive, phased
regulation can balance compliance with industry development.

The EU could benefit from a phased, flexible regulatory approach

that allows data centers to adopt sustainable practices without
compromising operational growth. Prioritizing real reductions in
energy and resource consumption would help shift the focus from mere
compliance toward meaningful operational improvements. As compliance
standards evolve, a framework that promotes innovation and keeps
sustainability at the forefront will be vital for the data center industry.

Whilst the EU reporting framework is harmonized, individual countries

are allowed to implement it in their own way. This could have negative
consequences on achieving the stringent targets set by governments.

29
 SUSTAINABILITY REPORTING AND EU COMPLIANCE CHALLENGES

B) Outdated equipment that is using unnecessary resources

Another issue for EED compliance lies in the complexities of managing assets
in data centers, especially within colocation models that serve multiple
tenants. In these facilities, those responsible for reporting must gather
energy usage data across diverse client-owned equipment. However, many
tenants lack visibility into their hardware’s energy impact, limiting the
accuracy of total facility reporting.

Adding to this complexity is the prevalence of outdated or unused

equipment that remains operational, consuming energy without serving any
functional purpose. Some facilities even house equipment dating back to the
1980s, unnecessarily inflating energy consumption. These continue to drain
resources and add to the facility’s environmental footprint.

“The reporting of power consumption from utilities and the building is fairly
straightforward, but the building is not the problem,” noted Mark Acton.
“It is the IT stack and networking that the industry must get a handle on.
Until there is appropriate awareness and metering in place, this is unlikely to
happen. Remember that in a well-run data center 80% or more of the power
consumption is driven by IT equipment demand, not the building power and
cooling infrastructure.”

C) Are silos a cause for concern?

Organizational silos present additional challenges, as IT and facility teams

within data centers often operate independently. This separation can
impact asset management, obscuring opportunities to optimize and
streamline energy use. Breaking down these silos through improved cross-
functional communication is essential to achieving operational transparency.
Data centers can better align their energy management strategies by
ensuring all stakeholders maintain a unified view of operations.

Colocation providers can assist clients in identifying and retiring

outdated equipment, which will help streamline reporting, reduce energy
consumption, and enhance overall operational efficiency. Comprehensive
transparency tools that provide insights into each asset’s energy usage
empower clients to make informed decisions about asset management
and support their sustainability goals.

LEVERAGING ADVANCED Advanced DCIM tools can support data centers in achieving reporting
SOFTWARE TOOLS compliance and, in some instances, may even help create a standardized
framework for ESG reporting. Advanced DCIM tools can give visibility
FOR REPORTING
into all assets—regardless of location or ownership—supporting accurate
and compliant energy reporting. DCIM tools like XpedITe enhance these
capabilities by automating data collection and reporting, reducing manual
intervention, and minimizing the risk of human error. Advanced tools like
this will provide a robust foundation for meeting regulatory requirements
while driving genuine energy efficiency and sustainability improvements.

30
SUSTAINABILITY REPORTING AND EU COMPLIANCE CHALLENGES

ADDRESSING ENERGY The increasing demand for Artificial Intelligence and advanced technologies
CONSUMPTION: has intensified concerns around energy consumption, especially as data
centers face mounting pressures on local and national power grids. With
CHALLENGES AND
compute requirements steadily growing, the environmental impact of data
MISCONCEPTIONS centers is gaining attention from regulatory bodies and communities near
these facilities.

A fundamental misconception in this context is the confusion between power

capacity and actual energy usage. Power capacity—the maximum power a
data center can draw—is often mistaken for actual energy consumption,
leading to misunderstandings around a data center’s environmental impact.
This confusion can result in exaggerated or misleading energy usage metrics,
contributing to inaccurate reporting and inflated consumption claims.

Power Usage Effectiveness (PUE) remains the prevailing metric and is widely
adopted because it provides a simple, standardized way to assess how
effectively a data center utilizes power. However, it will not give an overall
accurate picture of the energy usage as it does not include the IT stack. Also,
in some instances, data center operators apply their own way of calculating
‘PUE’ rather than using the global ISO standard, causing concern about
the true picture. As scrutiny from regulators and clients increases, there
is a growing call for independent verification of these metrics. Third-party
auditors could enhance transparency in reporting, ensuring that efficiency
claims are accurate and credible.

The increasing adoption and referencing of the ISO 30134 series of standards
marks a significant step toward global standardization in sustainability
metrics. By evaluating data centers across eight metrics, from Water Usage
Effectiveness (WUE) to Carbon Usage Effectiveness (CUE), the industry gains
a more holistic view of environmental impact. Implementing these metrics
encourages a shift from narrow efficiency benchmarks to broader, more
meaningful sustainability improvements, moving beyond the limitations of PUE
as a sole (and inappropriate) proxy for efficiency.

STANDARDIZING Although the EU has provided guidelines for efficiency in data centers,
BEST PRACTICES FOR uptake remains limited. Many data centers continue to operate based
on legacy processes that do not align with modern sustainability goals.
ENERGY EFFICIENCY
Organizational inertia, or resistance to changing established workflows,
remains a barrier to adopting these best practices. For example, colocation
operators could facilitate energy efficiency by offering optimization
services during client onboarding. However, there is skepticism about
whether clients will engage.

To encourage the adoption of best practices, data centers may need to

enhance client education, emphasizing the long-term operational and
financial benefits of efficient practices. Partnerships between data centers

31
 SUSTAINABILITY REPORTING AND EU COMPLIANCE CHALLENGES

and their clients can play a crucial role in realizing these improvements,
with EU-backed incentives further supporting the alignment of client and
data center sustainability objectives. Data centers can also set internal
goals for best-practice adoption, measuring their operational efficiency
and regulatory compliance progress.

ADDRESSING LIMITATIONS A recurring theme in sustainability reporting is the need for reliable data
IN REPORTING TOOLS collection and reporting tools. Many data centers use systems that lack
precision, with poor metering inside the facility and legacy software
AND METRICS
undermining reporting accuracy. These limitations prevent data centers
from obtaining precise measurements of energy usage, emissions, and
other environmental metrics, complicating compliance efforts and
diminishing trust in reported figures.

The industry will benefit from the creation of standardized, certified

reporting tools and processes. Independent consultants or auditors could
assist data centers in ensuring that reporting tools meet compliance
standards and validate the accuracy of collected data. As reporting
requirements grow more rigorous, these improvements will help data
centers confidently demonstrate compliance while providing transparency
to stakeholders.

Advanced Data Center Infrastructure Management (DCIM) tools

and Universal Intelligent Infrastructure Management (UIIM) systems
offer substantial benefits for meeting EU reporting requirements by
streamlining, automating, and enhancing the accuracy of sustainability and
energy reporting across data centers. Here’s how they specifically support
compliance and reporting for EU regulations like the Energy Efficiency
Directive (EED) and Corporate Sustainability Reporting Directive (CSRD):

i) C
 ENTRALIZED DATA Advanced DCIM tools and the Universal Intelligent Infrastructure
COLLECTION AND REAL-TIME Management (UIIM) methodologies can centralize the collection of
MONITORING diverse data streams across the facility, including the IT stack, energy
consumption, water usage, and emissions. This can also be done from
multiple data centers. These tools ensure consistent reporting metrics
aligned with EU-wide standards by integrating data across all facility
locations and systems.

Real-time monitoring provides up-to-date information on energy usage

and other sustainability metrics, essential for regulatory compliance and
internal decision-making. This live data can help operators quickly identify
inefficiencies and take corrective actions, aligning with EU goals for
ongoing energy performance improvements.

32
SUSTAINABILITY REPORTING AND EU COMPLIANCE CHALLENGES

ii) AUTOMATED REPORTING These systems can automate the generation of reports required by EU
AND STANDARDIZATION regulations, reducing the risk of human error and making reporting more
efficient. By standardizing data formats across all facilities, these systems
support uniform reporting that aligns with EED, CSRD and other ESG
reporting requirements. Automated reporting also makes adopting future
metrics and requirements easier as they evolve, enabling data centers
to remain agile and responsive to regulatory changes without requiring
extensive manual adjustments.

iii) ENHANCED VISIBILITY INTO By providing granular visibility down to individual asset performance,
ASSET LEVEL PERFORMANCE particularly useful for data centers operating under models where asset
ownership and accountability are shared, the operator can capture detailed
information on each asset’s energy usage and environmental impact -
compiling accurate data, even when client-owned assets are part of the
equation. With detailed asset data, operators can more accurately report
and justify their environmental impact, demonstrating compliance while
identifying under performing assets that may need replacement.

iv) SUSTAINABILITY INSIGHTS Advanced DCIM tools often include sustainability reporting modules
AND RENEWABLE ENERGY that track renewable energy usage, greenhouse gas emissions, and other
TRACKING ESG related metrics, aligning directly with CSRD and EED reporting
requirements. By tracking renewable energy contributions, these tools help
data centers demonstrate their commitment to EU climate goals. Detailed
insights on renewable energy sources and other sustainable practices allow
operators to quantify and showcase their investments in clean energy,
enhancing transparency and potentially improving public perception.

v) SUPPORT FOR TRANSPARENT, EU regulations increasingly demand compliance and verifiable and
AUDITABLE REPORTING transparent reporting practices. Advanced DCIM can provide audit
trails and secure data storage, enabling data centers to offer proof of
compliance through easily retrievable historical data. This transparency
is crucial for regulatory bodies and clients, ensuring the accuracy
and reliability of sustainability claims, thereby establishing trust and
accountability.

33
 SUSTAINABILITY REPORTING AND EU COMPLIANCE CHALLENGES

ROUNDTABLE Standardize ESG implementation:

RECOMMENDATIONS Address a fragmented approach to implementation by advocating
for a standardized approach, EU-wide.

Leverage advanced tools:

Adopt advanced DCIM tools within the UIIM framework for centralized
data collection, real-time monitoring, and automated reporting to
streamline adherence to EED and CSRD directives.

Enhance asset visibility:

Centralize tools that can monitor tenant-owned assets but retaining
confidentiality, ensuring comprehensive data collection and compliance.

Manage Legacy and Outdated Hardware:

Identify and decommission outdated or unused equipment to reduce
energy waste and environmental impact.

Encourage collaboration:
Break down silos between IT and facility teams to improve transparency
and energy management.

Ensure auditability:
Provide transparent, verifiable data trails to enhance credibility with
regulators and stakeholders.

Incorporate third party verification:

Engage independent auditors to validate energy performance claims.

Incentivize sustainability practices:

Collaborate with tenants on energy optimization during onboarding
and operations.

Align with long-term business goals:

Focus on TCO Reduction: Shift the narrative from short-term cost savings
to long-term resilience and operational efficiency.

34
SUSTAINABILITY REPORTING AND EU COMPLIANCE CHALLENGES

CONCLUSION: The rising emphasis of legislation such as sustainability in the data

EMBRACING SUSTAINABLE center industry underscores the importance of creating frameworks and
systems for accurate, comprehensive, and transparent reporting. With
PRACTICES IN
EU regulations like the Energy Efficiency Directive (EED) and Corporate
DATA CENTERS Sustainability Reporting Directive (CSRD) driving new benchmarks, data
centers face the dual challenge of meeting rigorous reporting requirements
while navigating diverse, potentially fragmented national regulations. In
addition, beyond compliance, robust sustainability reporting can serve
as a strategic advantage, demonstrating data centers’ commitment to
environmental stewardship and helping to align industry practices with
global climate goals.

Adopting advanced Data Center Infrastructure Management (DCIM) such

as Universal Intelligent Infrastructure Management (UIIM) is pivotal to
achieving this. Advanced data collation and management tools enable data
centers to capture real-time insights into all critical drivers in the data
center including energy use, emissions, water consumption. Automating
data collection and standardizing reporting processes will enable adherence
to regulation and build operational efficiency.

In the broader context, standardized reporting frameworks could further

enhance compliance, fostering EU-wide alignment on sustainability metrics.
As data centers increasingly leverage advanced infrastructure management
tools, they meet regulatory expectations and position themselves as
leaders in sustainable innovation. Ultimately, embracing this shift towards
transparent, responsible reporting is essential for ensuring long-term
resilience and creating an industry-wide orientation toward greater
sustainability.

35
 APPENDIX 1: ROUNDTABLE LEADERS

ABOUT THE AUTHORS

MARK ACTON Mark Acton is a seasoned expert with over 25 years in data center operations,
BUSINESS STRATEGY & TECHNOLOGY renowned for ensuring 24x365 availability in world-class facilities. His
DIRECTOR, RiT Tech
specialisation in delivering business-critical services is matched by a respected
leadership role in the data centres sector. Mark’s career is distinguished by his
international experience and a robust skill set that spans data centre facilities
design, and the combined management of IT and operational facilities.

Known for his technical acumen in consulting, Mark excels in integrating

sophisticated design with pragmatic management, setting new
benchmarks in operational excellence. His global perspective enriches his
approach, making him a versatile and insightful leader in the field.

Mark’s contributions have consistently advanced the standards of

data centre operations, earning him recognition as a visionary in
the industry. His dedication and innovative strategies continue to
inspire and shape the future of data centre management.

JEFF SAFOVICH Jeff brings over 25 years of substantial experience in the technology sector,
CTO, RiT Tech distinguished by significant contributions to tech product development and
innovation. His entrepreneurial journey includes the co-founding of three tech
start-ups, notably SphereUp, which achieved successful acquisition, and Zoomd,
which notably went public in 2019. His tenure at Comverse,
where he led a development group, further exemplifies his
expertise in technology leadership and development.

Jeff has been instrumental in revolutionizing Data Center Infrastructure

Management (DCIM). His pioneering introduction of the Universal
Intelligent Infrastructure Management (UIIM) concept has marked a
significant advancement in the industry. This innovation under Jeff’s
guidance has redefined the management of data centre infrastructures,
setting new standards for efficiency and intelligence in operations.

Jeff’s profound understanding of technology, coupled with his strategic leadership,

has not only propelled RiT Tech to new heights but has also significantly influenced
the evolution of data center management practices. His contributions to the field
reflect a deep commitment to excellence and innovation in the data centre industry.

36
APPENDIX 1: ROUNDTABLE LEADERS

ANTONIO SUAREZ Antonio Garcia, an eminent XpedITe Global Product Manager at RiT Tech,
GLOBAL PRODUCT MANAGER, is a distinguished expert in Data Centre Infrastructure Management
RiT Tech
(DCIM) and an advocate for Universal Intelligent Infrastructure
Management (UIIM). His career is fuelled by a passion for creating
accessible, impactful technology solutions in the data centre sector.

Holding a degree in Industrial Electronics from the Universidad Carlos III de

Madrid, Antonio has honed his expertise in developing and promoting effective
DCIM tools. His pivotal role at RiT Tech involves enhancing XpedITe, ensuring it
aligns with customer needs in both European and US markets. This task reflects
his dedication to evolving DCIM solutions in response to real-world demands.

Antonio’s approach is grounded in practicality and innovation, aiming

to make complex technology user-friendly and efficient. His work
not only advances XpedITe but also contributes significantly to the
broader conversation on intelligent data centre management.

Antonio brings extensive knowledge and experience, offering

insights into the future of data centres, shaped by intelligent
infrastructure and customer-centric design.

SUSAN ANDERTON Susan Anderton brings a wealth of knowledge and understanding of

MARKETING LEAD & EDITOR writing for the data center industry. Her 30 years experience as a brand
RiT Tech
consultant, writer and marketer results in an excellent understanding
of brand positioning and persuasive writing for industries. For over five
years, Susan has been gaining an in-depth understanding of the sector,
specifically in the DCIM arena, construction and professional services.

As RiT Tech’s Marketing Lead, Susan has been instrumental in evolving marketing
strategies to meet the distinct challenges of DCIM. She is renowned for her
ability to clearly communicate on complex data center technologies and trends.

37
 APPENDIX 2: DELEGATES AND PARTICIPANTS

COMPANIES THAT
ATTENDED SEPTEMBER 24 EVENT

Enterprise Data Centers

State Street Bank

Goldman Sachs

Technology

NVIDIA

Analysts

Uptime Institute

Supply Chain

EnerSys

JLL

WSP UK

Qcloud

Kevlinx

SNHA Woolpert

Andget Ltd

George Fischer

Korgi Consulting

APK Group

AtkinsRéalis

Onnec

MCW

Viadex Global

Durata

Media

DCD>Academy

38
Copyright RiT Tech (Intelligence Solutions) Ltd. 2023 | Company No. 515565216