Digital Transformation in Water Utilities: Status, Challenges, and Prospects

Highlights

What are the main findings?

• Digital transformation in water utilities occurs along two pathways, one to serve business operations and the other to operate and maintain their intensive infrastructures with evolving tools like data-centric asset management, decision support systems, SCADA systems, digital twins, and process optimization.
• Digital advances in water utilities occur more through their support community of their own technology gatekeepers, consultants, and vendors than through academic and government research projects, and technological innovation will evolve slowly among leading edge utilities, as many others are too stressed with basic issues to adopt many of them.

What is the implication of the main finding?

• Support for utility leadership in data-centric management offers a promising avenue to overcome legacy problems of aging infrastructure, water losses, and contamination problems in drinking water networks.
• Promoting innovation through utility networking and peer-to-peer knowledge transfer offers policy leaders a promising pathway to improve urban water services and requires understanding of the importance of knowledge management in allocating financial resources for utility work.

Abstract

While digital transformation in e-commerce receives the most publicity, applications in energy and water utilities have been ongoing for decades. Using a methodology based on a systematic review, the paper offers a model of how it occurs in water utilities, reviews experiences from the field, and derives lessons learned to create a road map for future research and implementation. Innovation in water utilities occurs more in the field than through organized research, and utilities share their experiences globally through networks such as water associations, focus groups, and media outlets. Their digital transformation journeys are evident in business practices, operations, and asset management, including methods like decision support systems, SCADA systems, digital twins, and process optimization. Meanwhile, they operate traditional regulated services while being challenged by issues like aging infrastructure and workforce capacity. They operate complex and expensive distribution systems that require grafting of new controls onto older systems with vulnerable components. Digital transformation in utilities is driven by return on investment and regulatory and workforce constraints and leads to cautious adoption of innovative methods unless required by external pressures. Utility adoption occurs gradually as digital tools help utilities to leverage system data for maintenance management, system renewal, and water loss control. Digital twins offer the advantages of enterprise data, decision support, and simulation models and can support distribution system optimization by integrating advanced metering infrastructure devices and water loss control through more granular pressure control. Models to anticipate water main breaks can also be included. With such advances, concerns about cyber security will grow. The lessons learned from the review indicate that research and development for new digital tools will continue, but utility adoption will continue to evolve slowly, even as many utilities globally are too stressed with difficult issues to adopt them. Rather than rely on government and academics for research support, utilities will need help from their support community of regulators, consultants, vendors, and all researchers to navigate the pathways that lie ahead.

1. Introduction

As artificial intelligence (AI) and machine learning (ML) tools advance, utilities and other business and service organizations are emphasizing digital transformation (DT) to exploit their possibilities [1]. The most familiar applications of DT are in e-commerce and interaction with chatbots, online services, and other digital interfaces, but in asset-intensive organizations like energy and water utilities, the focus and requirements are different [2]. Like many business organizations, these utilities use digital tools for consumer interactions, but their DT programs reach much deeper into the management of their extensive and asset-heavy infrastructure systems. Given their many challenges, DT offers a path to improved services and effectiveness, but how it is transforming them is not well understood amidst the extensive coverage of new technologies like AI and ML. This paper focuses on water utilities to trace how DT applications have evolved over time and to provide a framework to explain the status and prospects for new DT advances in them. Experiences among water utilities show a lack of awareness among utility managers and researchers about the actual advances and possibilities provided by DT. The paper identifies the specific technologies and methods used for DT in utilities and it probes how these are being implemented, with an emphasis on actual rather than claimed applications. It assesses the research base from the perspective of return on investment and the incentives of vendors and developers and it offers a road map toward the future for advances in research and implementation of DT in water utilities.

One driver for the work is the work of the Smart Water Task Force of the International Water Resources Association (IWRA) [3]. The work of the Task Force and its case studies indicated many diverse and situation-specific topics about smart water and smart cities, but they did not provide much insight into DT applications in water utilities. Another driver of the work reported here is the writer’s own experiences with utilities and water sector research organizations, which has provided insight into how advances are studied and adopted and barriers to innovation in the highly regulated asset-intensive utilities.

2. Methodology

The methodology for the review (Figure 1) builds on data from the writer’s survey of drinking water utilities reported in a previous paper [4] and leads to a further study of technologies, methods, and implementation of specific DT strategies. Like other emerging technological developments, DT lacks a standard definition and interpretation for specific industries. This barrier is confronted by defining a concept for DT and then distinguishing its applications among the general case for all businesses and utilities. This is followed by parsing its applications according to the operational and business processes of utilities, which leads to a water utility DT model.

The water utility DT model is based on the archetype of a standard utility organization, which is an important step in the methodology because, without focusing on the most important advances in DT applied to distribution system infrastructure, its use for business operations like financial management is not significantly different from other subscription-type business models. The next step in the methodology is to investigate how DT has evolved as applied to distribution systems.

The literature reviewed stems mainly from utility industry sources, which report more rapidly about fast-moving innovations than the academic literature. Archival research literature can be misleading about the status of implementation of innovations because investigators are incentivized to report novel advances, but proof of application and results may be lacking. For example, the use case cited in the paper of water loss control shows how DT for this important goal stems back more than three decades and does not depend on current research publications about technological developments like AI and ML.

The use cases explained in the review of DT evolution provide a longitudinal view of how it has evolved and is understood within water utilities. Only with specific examples like this can one follow the somewhat fragmented and stepwise progress of introducing and testing new methods, which sometimes endure and in other cases do not.

The “Lessons Learned” section leads to a roadmap for utilities, including recommendations for research as well as organizational development. The concept is to focus on promising but practical applications of DT that fit within the set of incentives and constraints facing utilities, as well as to address the challenges faced by utilities that lack financial and management capacity to innovate.

3. Digital Transformation in Water Utilities

The concept of DT is familiar because everyone experiences it in everyday life. From web-based shopping to dialogues with chatbots, people see a shift from interactions with people to encounters with computing devices, especially in e-commerce [1]. Business scholars explain that DT is about rewiring how organizations operate, deploying technology to respond to customers, and lowering costs [5]. Some discussions are about how DT impacts business organizations, such as requiring more educated workers, leading to fewer low-wage jobs and creating anxiety in workforce operations like hiring, supervision, and automation [6,7]. Studies also show that its deployment includes organizational capacity building, digitizing operations, and marketing [8].

These deployments of DT can be mapped from businesses to utilities to include business operations, system operations, and maintenance work. Most utility attention has been on electric power utilities [9], which offer a dynamic and rapidly changing sector for DT and where emerging technologies are disrupting business models with renewable energy, smart grids, and distributed sources like solar and battery storage. Application areas like electric vehicles, smart buildings, and data centers are changing demands, while end users still focus on cost and reliability, transparency, and cleaner energy. Like water utilities, energy utilities prioritize reliability while responding to customer choices about resource sources, environmental impact, and cost. They are monopolies, but they must respond to competition, such as EVs providing power and third parties monitoring electricity use. They are also challenged by legacy infrastructures and an aging workforce that may resist change or lack needed digital skills [10].

Water utilities share some attributes with electric power utilities and face similar challenges. Both are traditional services subject to regulation and operate in stable environments with some outdated technology and legacy systems. Both seek to integrate digital technologies into operations and customer relations while being challenged by the need to equip and prepare field workers with digital tools. They are also challenged by competition, trust issues, and consumer choices. Divergence in their operating environments focuses on the difference between the business worlds of energy and water, the direct connection of water with public health, and the unique attributes of water as a ubiquitous resource that touches people in many ways.

To illustrate levels of interest among water utilities, the American Water Works Association (AWWA) issued a series of “Water 2050” strategy documents and made recommendations about how to leverage technology to address barriers to innovation, regulatory compliance, and unintended consequences [11]. As with many such broad visions, the outcomes reflect general issues rather than specific and immediate needs, and the next step is to explore actual use cases to identify where change is happening.

A reality check shows that only a subset of all drinking water systems is prepared to adopt new digital technologies on a significant scale because much of the global population lacks access to organized utilities. The extent of such access is not known exactly as global statistics record the percentage of people having “piped water on premises,” but this does not indicate utility service, especially in rural areas. Such piped water can be problematic with low service levels [12].

Using available data, the writer estimates that the number of utilities ready for significant DT globally is on the order of 20,000, mostly in well-organized cities. In cities of high-income countries, most people are served by such utilities, although not all can afford the service [13]. For example, in the US, the number of customers served in urban areas with populations of 5000 or greater was about 270 million as of 1 January 2025 [14]. Drawing from statistics of community water systems, this indicates that about 6000 utilities in the US serve populations of 5000 or greater where most people live [15]. The situations in Europe and Japan are similar. Most water supply in England and Wales is provided by 32 privately-owned companies [16]. In Germany, there were approximately 6065 water supply enterprises and utilities in 2010, mainly small municipal utilities [17]. Data from the French National Agency for Water and Aquatic Environments [18] show that in 2009 there were some 14,217 drinking water services in France, but not all were well-organized utilities. Japan maintains a classification system and statistics for water supply services through its Ministry of Health, Labor and Welfare [19], and ownership categories for “Water Supply Business” showed 1572 utilities in 2017, mainly managed by municipalities.

The estimate of on the order of 20,000 utilities globally is meant to identify a set of them with adequate organization and finances to undertake DT on a systematic basis. These utilities generally have extensive distribution systems with the possibility of exploiting the benefits of DT [20] for their legacy buried pipe networks and new components for growth, improvement, and control. These distribution systems are complex and expensive, and they must be sustained and managed, despite many challenges. This means that new controls and management systems must be grafted onto older systems, many of which are in impaired condition with leaks and vulnerable components.

4. Model of Digital Transformation in Water Utilities

The organizational arrangements within water utilities change slowly and this enables analysts to chart how DT aligns with business practices. To illustrate, Figure 2 displays a typical organizational chart and where the pillars of DT will focus in each operational category. DT can be divided into management capacity, operations, and customer relations categories [8]. Management capacity can be divided further into business practices and management of infrastructure. The business processes are like those of other organizational types, mainly human resources, finance, and marketing. Customer relations stem from business processes, especially the financial part. Maintenance of system infrastructure requires preventive and corrective maintenance as well as overall asset management. The operations category addresses methods like decision support systems, supervisory control and data acquisition (SCADA), digital twins, and related processes.

In a conceptual view, the way that utilities are advancing through DT for operations can be summarized as shown in

Table 1. Digital transformation of distribution systems.

	Physical	Hydraulic	Water Quality
Traditional	Reactive control	Hydraulic analysis	Basic monitoring
Transition	Active control	Basic models	New sensors
Digital	Data-centric management	Integrated models	Sensor networks

, which shows management for physical, hydraulic, and water quality objectives. Evidence shows that utilities began to take more interest in active management of distribution systems about three decades ago, and this enabled breakthroughs like movement from reactive maintenance systems, from calculators to computer models, and from basic monitoring to more advanced systems based on new sensors. The DT era takes the transition further, with data-centric asset management, integrated models, and newer networks of sensors distributed more densely in distribution systems.

5. Innovation in Water Utilities

Leaders in water supply utilities recognize the need for innovation to improve management effectiveness [21]. To assess progress, performance management can be used with metrics such as efficiency, compliance, service levels, and affordability [22]. However, performance management and benchmarking programs are not widely used among water utilities. To some extent, this is due to the conservative nature of water utilities, and it means that new paths to innovation and transfer to practice are needed.

Barriers are formed by ownership of most US water utilities by local governments, along with state and federal regulation for health effects, but regulation only by local political boards for rates and overall performance. These boards often lack expertise and may also lack incentives to implement management improvements. Privately owned utilities are indirectly regulated for management effectiveness as they must justify rate increases to regulatory commissions. Another challenge is that while progressive and well-funded water utilities may implement research findings, transferring them to other utilities faces more challenges.

Most research projects offer only modest advances, and results are rarely implemented on a large scale. Researchers tend to focus on the development of tools that can be published in journals, but enduring innovation occurs over time and requires the involvement of proactive managers, in-house experts, and industry vendors. It can occur by the actions of technology leaders working through water industry networks when groups with a common purpose are networked [23].

As an example, the innovation of a new method to audit water losses is an object lesson for innovation. The process started around 1990 when a vision for a comprehensive approach to water loss control emerged among leading water associations. Work was organized via committees and task forces of the International Water Association (IWA), which was formed in 1999 as a merger of the International Water Supply Association (IWSA) and the International Association of Water Pollution Research. AWWA leaders participated in the work and released the first edition of Manual of Practice M36 on Water Audits and Leak Detection in 1991.

Long-term leadership was exercised by Allan Lambert, a British engineer who chaired the first IWA Water Loss Task Force. He had water utility experience and was prepared due to work done beginning in the 1980s, when measuring leakage was important in the privatization of UK water utilities. The work of the IWA Water Loss Task Force, comprising 22 national reports, is summarized in its 2000 report [24], which documents the collaboration beginning in 1991 with the IWSA Report on “Unaccounted for Water and the Economics of Leak Detection.”

Now, thousands of utilities use the method, and some state agencies have begun to require it. The innovation was explained this way: “Today’s budding water loss industry grew out of the efforts of a bunch of brilliant, obsessive, far-thinking engineers in Britain who started something called the National Leakage Initiative in the early 1990s.” “Led by a man named Allan Lambert, they developed a methodology for categorizing and quantifying water leakage, and predicting losses, so they could rigorously determine how to reduce them. This was vital in Britain, which had some of the world’s oldest water systems.” “Their efforts were famously successful” [25].

Regulatory pressure helped drive the innovation and, without it, voluntary adoption of new methods may not have occurred. In the absence of regulatory pressure, other research-based innovations, like pipe break models and distribution system optimization, have not taken off, despite considerable research on them. Another driver is return on investment, which can enable utilities to keep rates low and avoid political pressure.

Ultimately, innovation will require proactive managers, technology entrepreneurs, and industry vendors more than planned transfer from research projects. Despite this experience, dialogue continues about improving the research-to-practice process [26]. These discussions usually focus on research-based tools like models and optimization algorithms and participants in the dialogue tend to be consultants, academics, or technical gate keepers in utilities.

6. Chronology of Use Cases of DT

In a general sense, there is much discussion about DT because it applies to organizations across the spectrum of industries and businesses. DT studies seem mostly focused on companies that build products and services for end consumers, but in asset-intensive industries they face different strategic and organizational challenges [2]. Water distribution systems are asset-intensive because sustaining the infrastructure is the dominant management issue [27], along with cyber security.

Progress in DT of drinking water utilities has been occurring for decades. For example, research papers go back at least to 1957 on computer-based distribution system analysis [28]. In the 1960s, Murray McPherson, director of the Urban Water Research Program (UWRRP) of the American Society of Civil Engineers, was working on automatic control of distribution systems, having gained experience as a “water and sewer research engineer” at the Philadelphia Water Department [29]. As part of the UWWRP, he led extensive research on innovation using computers and information systems [30]. The writer worked with McPherson in the 1970s on a study of automatic control of combined sewer systems and it was evident by then that SCADA systems were evolving as operational technologies [31] for use in distribution system network models and system control strategies. Utilities also embraced GIS and advanced databases as they emerged, often with support from innovative vendors [32].

To provide perspective on the evolutionary implementation of DT by water utilities,

Table 2. Chronology of implementation of digital transformation elements.

Inception	Technology	Description
1950	SCADA systems	SCADA developed along with communication and monitoring methods as it became possible to report system behavior and needs remotely
1957	Computer simulation of distribution systems	Research and software commercialization led to maturation of distribution system models by about 2000
1960	GIS	Computer-based mapping developed rapidly along with computer capabilities and new methods of image processing
1980	Enterprise databases	Database management entered widespread use during the 1980s and 1990s as major software vendors developed new packages
2000	Asset management systems	Emergence of GIS and data management enabled new processes for asset management and adoption and standardization across industries
2000	Water loss control audits	Utility networking to address water losses began around 1990 and by 2000 the collective work had developed a standardized method
2015	Advanced metering infrastructure (AMI)	AMI emerged after about 2010 to assemble arrays of sensors and communications to address various network issues in electrical and water utilities

lists the approximate dates when key technologies became operational. These dates are approximate because technologies usually develop incrementally. Each technology has its own history, with many players and events. Technologies like the Internet and software packages for general use are not listed.

Recent visioning among AWWA leaders aims to extend and advance this past work [33,34]. Also, studies sponsored by the WRF confirm the interest and rapid movement of technological applications related to DT [35]. The proof of its success lies in the use cases for applications across the pillars of DT.

6.1. Organizational Capacity Building

Management capacity across the organization is a leading concern of utilities seeking to apply DT technologies [36]. Their business services are like those of other public and private sector organizations, although public sector organizations may face more constraints in personnel management and purchasing due to government rules [37]. Otherwise, employee data, hiring, retirements, benefits, and other human resources functions will be similar. On the financial side, utility billing and customer information will be similar in water and energy utilities, but differ from the private sector focus on profitability, sales, and market growth. Many changes in the application of AI to utility billing are apparent, and links with customer interaction goals are increasing to respond to consumer desires [38].

Organizational capacity to manage the condition and performance of system infrastructure is a unique challenge facing asset-intensive distribution systems. To define the problem, generalized data began to appear shortly after passage of the Safe Drinking Water Act in 1974, and this fueled interest in the use of data to manage infrastructure inventory, maintenance scheduling, failures, and renewal. To respond to this need, the method of asset management has been growing in use and priority among utilities [39].

The collection and use of utility data for management purposes have evolved rapidly within the water supply industry. AWWA’s surveys to provide infrastructure data began before the emergence of computers. During the 1980s the efforts intensified, and the association is now collecting utility data for benchmarking systems [40] and performance assessment [22]. This coincided with a paradigm shift from passive to more active management of distribution systems that occurred during the 1980s. In parallel, the US Environmental Protection Agency began to survey community water systems, with a focus on infrastructure inventory, condition, and financial needs. Surveys were conducted in 1995, 2000, and 2006 [41].

As data systems emerged, the WRF was pioneering standardized formats for utility data management. A project to identify key asset data for water sector utilities recommended asset hierarchies for use in data identification and management, asset hierarchies, attributes, and performance indicators [42]. However, as with other research products, implementation by water utilities has been slow.

While much work has been done to collect utility data and use it to improve management [43,44], most applications are supported by commercial software packages, which have proved more attractive to utilities than in-house approaches [45]. Some software packages, such as distribution system models, have proved more popular than others. For example, the writer participated in a project to develop software to rate risks of main breaks and it proved useful, but utilities lacked workforce capacity and incentives to use it [46].

As computer systems improved in the 1970s and 1980s, water utilities focused on digitizing legacy data from work orders, maps, and other technical information sources. The writer witnessed this in two large utilities, one in the UK, which was preparing for privatization, and another in the US, where legacy infrastructure data were being captured in databases with numerical and anecdotal data, photographic copies of drawings, and related system data.

With the emergence of GIS for management of distribution systems, a different revolution has occurred. Now, spatial and non-spatial data are combined in one platform and facilitate visualizing and managing the spatial infrastructure in distribution systems [47,48]. In the writer’s survey mentioned earlier, practically every utility reported using a GIS to manage distribution systems.

In a general way, asset management is useful across different fields by offering new ways to perform tasks like keeping records and scheduling work [49]. Utility asset management uses a data intensive approach to decision making for maintenance, rehabilitation, renewal, or replacement of distribution system assets, as well as other infrastructure. The method began to evolve around 2000 and, in 2008, the WRF produced a research road map for it [39]. Despite its potential, progress in the implementation of asset management has been relatively slow, and there is no regulatory requirement to incentivize utilities to adopt it [50].

6.2. Distribution System Operations

In parallel with asset management, digitization of operations has become a a major topic of DT in utilities. At a general level, this refers to the activities shown in Figure 3, which illustrates a system to be operated (distribution system), data to be collected from sensors (like pressure, flow rates, water quality), decisions to be made (like needed additional flow or pressure regulation), and commands sent to actuators (such as valve operators or pump controls). As an example, a model can be used for tasks such as computation of water age, which might in turn inform the decision to change a flow rate, for example. The models and decision logic can be considered as a decision support system, and they are commonly associated with SCADA systems that organize data and provide commands to be sent.

SCADA systems began with phone and radio alerts and transitioned to use of computers as they became available [31]. The field of control engineering coalesced to organize the needed devices in the form of sensors, actuators, control panels, and related equipment. Now SCADA systems are ubiquitous and require ongoing support, such as that provided by AWWA by its operational technology committee [51].

Cyber security is a major issue in DT, especially for asset-intensive organizations, and SCADA systems provide a point of vulnerability. The disadvantage of networked information and control systems for critical infrastructure is that they are vulnerable to hacking. This is, of course, a general problem among all critical infrastructures and it worsens as systems become more networked and hackers become more sophisticated [52,53,54].

Digital twins are also emerging rapidly, but they are not uniform because diverse uses are being developed across many application areas [55]. They are related to traditional enterprise information systems and decision support systems, both of which were initiated some four decades ago [56,57]. In the water industry, discussion is occurring among a relatively small group of stakeholders, some who are organizing through AWWA [58] and the Smart Water Networks Forum (SWAN) [59]. Such attention by small groups of advocates is not unusual in the initial stages of technology development. Definitions by such groups can seem vague, like a digital twin is “A digital, dynamic system of real-world entities and their behaviors using models with static and dynamic data that enable insights and interactions to drive actionable and optimized outcomes” [60]. Unless you were in the group that developed that very general definition it would be hard to understand it. Like many general concepts, digital twins can be explained better through examples [61,62]. One such example was by Ostfeld et al. [55], who offered a concept of three layers, a GIS to organize basic data, a system interpretation layer with various models and computational capabilities, and a decision support layer.

The availability of advanced metering infrastructure (AMI) is enabling much of the DT in utilities. The term means a system to record the consumption of energy or water on an hourly or more frequent basis and to transmit measurements over a communication network to a collection point [63]. In electric power utilities, it is used to manage distribution lines, collect interval data, detect tampering and outages, study loads due to devices and premises, and for other tasks. Water utilities adopted the AMI term from energy utilities and can use the devices for tasks such as automated meter reading, studying water losses, managing peak loads, and planning for growth. Although the water industry has been slower than the electric power industry to embrace AMI, growing demand, shrinking supplies, higher operating costs, and workforce challenges make them appealing. Prospects include remote shut-off, real-time water quality monitoring, and pressure management capabilities [64].

Controlling water losses from distribution systems is a common problem in water utilities, and DT provides tools to help with the task. A promising strategy is time-of-day pressure control that combines pressure sensors with SCADA and algorithms to manage pumps and valves to control pressure ranges. Efforts began prior to 2000 to develop the digital data-based method for water loss auditing discussed earlier [65]. The research process for the method is a good example of DT advances in the water industry. The problem of massive water losses was recognized and, even by the 1950s, AWWA leaders were working on water loss control. The subsequent work was organized via committees and task forces of the International Water Association (IWA). Work of the IWA Water Loss Task Force led to widespread adoption of a standardized water loss audit methodology [60].

Water main failures are another ubiquitous problem affecting utilities, and with DT they can be anticipated, to some extent. Attempts have been made to model the risk of breaks, but analytical tools are not used by most utilities. The problems stem from data quality and complexity of statistical models. Tools based on machine learning may improve prospects for use of available data, and off-the-shelf software may nudge utilities to adopt the models more broadly [66].

Given the multiple demands placed on utility managers, the process of distribution system optimization (DSO) is appealing. It means to monitor and control all processes in an integrated way to meet performance goals. Water quality optimization requires the capability to simulate water age [67], and the concept of optimization has been extended to include the hydraulic and physical performance of distribution systems. DSO was adopted by AWWA’s Partnership for Safe Water (PSW) program, which co-sponsored a WRF project to develop the concept [68]. DSO uses pressure, main breaks, and chlorine residual as indicators of physical, hydraulic, and water quality integrity, and the program depends on numeric criteria and certified self-assessment checklists. Participation in the PSW program for distribution systems is low, likely because it is evolving and not required by regulators. In the writer’s survey, most utilities indicated a lack of understanding of it. Use of the term “optimization” may be problematic as it can mean different things, such as optimization of hydraulic network design, a popular research topic.

The use cases described represent areas of emphasis selected by utility leaders through their participation in the WRF or by publication in industry journals. In addition, papers are emerging that announce new tools in search of specific applications but lack proof-of-concept. Examples include how new machine learning tools may help with treatment plants, water reuse, and related control issues [69,70,71]. General utility issues like drought vulnerability can be studied with machine learning [72]. These papers offer the possibility for further research to investigate the feasibility of their promised applications.

6.3. Enterprise Management in Utilities

The use cases illustrate many instances where DT is affecting all enterprise activities, including operations, asset management, and interaction with customers. The resulting enterprise management systems (EMSs) are successors to the older concept of management information systems and can have different names, including enterprise content management, enterprise resources planning, among others.

Logically, the EMS can be used at organizational levels for different purposes, at the executive level for data to support strategic decision making, at the managerial level for planning and controlling organizational functions, and at the operational level for business process management. These uses span business processes like document management, customer relationship management, and human resources.

These business processes must consider both utility and customer perspectives. Revenue is a critical issue for utilities for which devices for metering infrastructure are essential. Customer interaction with the utility and especially with rates and service issues is increasingly important. Water utilities have many new technologies available for functionalities from management of accounts to support for customer self-service. Meter data management systems are emerging for use with data analytics to study customer segmentation, consumption and payment patterns, and location-based studies. Information systems are paired with GIS and maintenance management systems to provide utilities with integrated analysis capabilities. Customers increasingly use social media to manage their accounts online [73].

Such management systems appeal to many types of organizations and software companies seeking new market niches. Most work has been for private sector firms and among information science researchers, and little effort has been directed at the public sector, including utilities [74]. The concept of the EMS is more logical for utilities than for other public sector organizations because they have processes and customers that are like many other types of businesses. An example of such an EMS is at Pennsylvania State University, which uses it for its utilities [75]. Among water utilities, popular names for it, like EMS, are not used much, but attention is focused on DT and on data analytics. The WRF [76] sees it related to shifting careers for data professionals in the water sector.

While terms like EMS seem to come and go, a coalescence is apparent among concepts related to DT in utilities. This involves the merging of concepts for enterprise management, decision support systems, digital twins, simulation models, and centralized data support. Such a merger is conceptualized in different ways, such as in [77], where the authors combine data analytics and management, visualization, DSS, and models to provide user experiences through dashboards and communication links. That conceptualization uses the digital twin as the enterprise management framework, but it focuses on system operations and does not seem to extend to business processes that affect customers.

Figure 4 offers a comprehensive view using the enterprise management space as the boundary for activities related to decision support, models, and data management. The operational models shown in Figure 3 are the heart of it. DT runs through it as the various data and information sharing links operate to use data for management of system operations, system status, and customer relations. The digital twin concept is embedded at three levels in this framework:

• Digital twin of the physical system;
• Digital twin of the physical system and its control mechanism;
• Digital twin of the enterprise with data analytics and visualization through dashboards and communication links.

7. Lessons Learned and a DT Road Map for Water Utilities

The literature and use cases show how DT in water utilities is evolving incrementally as a gradual transformation driven by return on investment (ROI) and the demands of sustaining regulated water services in different settings. As water utilities innovate, their goals are constrained by the need to sustain reliable services without regulatory violations and the limits imposed by utility workforce challenges. The net result is a demonstrated reluctance to adopt innovative methods unless required by external pressures. Utilities need a road map based on performance improvement, return on investment, and their regulatory environments to serve as a practical guide to considering DT implementation.

An example lesson is that performance management through use of indicators, benchmarking, and the DSO concept is progressing slowly. The AWWA SOTWI report does not include performance improvement among the top ten concerns of utilities [27], although it is implicit in the financial categories. So, in the category of lessons, performance management drops off the list of priorities in favor of conservative approaches to avoid regulatory sanctions, public disfavor, and system failures.

The announcements about DT and related technologies like AI, ML, digital twins, and optimized control suggest a promising future is ahead, but the proof is in actual utility experiences rather than research or vendor announcements about possible applications. Examples of utility experiences can be offered by the categories of DT based on capacity building, operations, system management, and customer engagement.

DT in business support services is proving helpful, as it is in other types of organizations. Utility billing and customer information are used in similar ways in water and energy utilities, and their use responds to the shared need for financial sustainability and improved transparency for customers. Improved capacity for technical management using digital methods is also promising, with the application of data analytics, GIS, and digital twins helping some utilities to cope with workforce challenges.

Data systems and data analytics are important due to their usefulness in managing capital-intensive water utilities infrastructures, especially distribution systems. They are used extensively to assess infrastructure conditions and performance to support asset management and operational needs. While the WRF pioneered standardized formats for utility data management, utilities favor commercial packages due to their lack of workforce capacity for research, development, and experimentation. Meanwhile, researchers are integrating and synthesizing legacy data systems to develop new data analytics methods to interpret system behavior, but the extent to which utilities will adopt new measures is uncertain.

GIS has been a bright spot for digital management of distribution systems. The combination of spatial and system data in one place is proving useful across many utility operations, and utilities are able to locate their GIS systems centrally so as to leverage data and workforce capabilities to take advantage of system capabilities. The future for GIS seems assured across many urban management applications related to distribution systems, such as management of street systems and other utilities that are co-located with water pipes.

Asset management is another promising technology, which will rise in importance as systems continue to age and require more attention. The method is data intensive and aligns well with DT, but this challenges utilities to do better in collecting and managing their system data. To support asset management, utilities can model risk of breaks, but models have not been implemented extensively. Machine learning may offer new possibilities. Reluctance of utilities to implement asset management can be traced to organizational causes and will require team-building approaches to make improvements.

In the operation of distribution systems, SCADA is ubiquitous and requires support to keep systems functional and effective. Cyber security is a major issue with these control systems, as is keeping systems maintained and functional. This is a special problem for smaller utilities with workforce and financial challenges. New sensors, actuators, and even IoT additions to systems will challenge the complexity of SCADA systems and the related digital accessories like display boards, performance checks, alarms, and related operational features.

Among the new technologies and methods, digital twins seem to offer substantial benefits and prospects. While there is some confusion about what they are, system models have advanced to a sophisticated level and continue to improve. Their future seems to offer an organizing platform for assembly of models, data systems, and control features for distribution systems management.

On the management front, DSO is important but has not been implemented extensively. Water quality optimization via simulation for water age has advanced, due to pressure from regulatory requirements and reporting. The other parts of DSO focused on computer-based performance management are promising, but they lack regulatory drivers and other proven ROI demonstrations, at least so far. The associated process of water loss control systems has been mandated in many US states, and they can pass the ROI test well in water-short areas. System management techniques like optimization of time-of-day pressure control offer possibilities for implementation of new sensors and actuators.

AMI offers promising applications, but systems have not been implemented extensively. Once again, the ROI test seems to explain why utilities are slow in implementing them. The key is in the limited funding utilities have for capital investment and infrastructure management. If the choice is about whether to replace risky system elements or to invest in experimental technologies, the risk reduction goal will dominate.

Most of the DT categories reviewed are methods rather than a device or management system. Among the latter, GIS and SCADA seem to have the greatest momentum and perceived value to utilities, while digital twins are still an evolving concept and not yet fully developed as a management system or set of equipment. AMI has received much attention, but rates of implementation suggest a “wait and see” attitude among utilities.

Among methods, business support and customer information systems are evolving rapidly, as they are in other industries. Water loss control offers return on investment and is subject to some regulatory requirements, so it appears to have a good future. Technical management methods offer much promise, but they seem to be most discretionary and lag somewhat in adoption. These include asset management, modeling of break risk, and DSO.

Innovation in DT occurs within a regulated environment where utilities keep a close eye on return on investment. It happens slowly and requires networking within the utility industry. Opportunities for researchers and vendors focused on promising areas not yet exploited fully. Business services and customer information systems have many competitors. The arena of technical management seems ripe for new products and services. Digital twin packages might be a promising market, especially if costs can be controlled. New devices such as AMI and related sensors offer good possibilities when they can pass the ROI test.

The use cases described offer promising pathways to elevate the performance of utilities, but water utilities undergoing DT must confront workforce challenges. Studies indicate that utilities need job descriptions for digital expertise and to attract learners who can develop new skills for use of technologies such as digital twins, drones, and virtual reality [78]. This is in addition to legacy workforce issues that have been created as utility workforces age, and many institutional memories fade away. These must be replaced by newer digital memories and the application of knowledge management programs [79,80] to identify key organizational knowledge, capture it when needed, and disseminate it to workers who need it.

A preface to identifying a road map for research must recognize why utilities resist implementing research findings. Due to their business models, they are naturally risk-averse, and unless required to implement new technologies or methods, they may not be willing to reach out to try them. The writer faced this experience when leading a project to implement new data technologies to minimize pipe breaks and found utility resistance due to workload constraints and lack of regulatory pressures or incentives to cooperate with other utilities.

Research to advance DT in water utilities occurs more among practitioners than in academia. One reason is that academic research tends to focus more on basic science and academics may lack access to real problems that utilities and their consultants and vendors work on. Another reason is the set of incentives for leading consultants and vendors to develop new products and services given that they serve research enterprises like the Water Research Foundation (WRF), which responds to the specific needs of its subscribers. This set of conditions led to two WRF projects to identify road maps for future research in DT arenas. The writer participated in both, one aimed at improving asset management and the other at generating a research agenda for a WRF focus area about “Intelligent Distribution Systems.” [81].

The research process for the intelligent distribution systems project involved experts from utilities, regulators, associations, academia, and innovative companies to identify topics and projects. It included a literature review for promising technologies for operational and customer-related arenas over topics like monitoring equipment, communication systems, data collection and analysis, regulatory changes, and elements in electric, gas, and oil distribution systems. The industry review focused on interviews at the annual AWWA conference as well as a review of industry literature, the Internet, and contact with vendors.

Examples of topics include: AMI, monitoring for water quality and pipe stresses, actuating valves, condition assessment, pipe location, and technologies like smart pipes, drones, pigs, and robots. Projects rated with the highest priority focused on near-term needs of utilities, like meeting power requirements, security and communication methodologies, using AMI to meet customer needs, assessing benefits, data integration and analytics, and use of DT for water loss control.

Future research needs for DT in water utilities can be divided into two lines, one for near-term utility uses like the projects described and the other for more basic examinations of the same topics. The basic topics are likely to focus on data analytics, cybersecurity, optimization methods using AI and ML, and new methods to interact with customers and the public. The literature review for the present study indicates a gap between current research on these topics and applications, but potential breakthroughs offer attractive possibilities for the future.

8. Conclusions

The review of water utility progress with DT shows that, despite the current trendiness of AI and ML, progress is apt to continue with gradual change and few breakthroughs. The new technologies are attractive and they offer a promising future, but experiences show that utilities adopt new technologies and methods by considering return on investment and their ongoing need to sustain effective water services in a regulatory climate with a range of socio-economic settings. Utility workforce challenges also show reluctance to adopt innovative methods unless required by external pressures or confronted with overwhelming cost-benefit data. These constraints seem to explain a general reluctance among utilities to adopt higher levels of performance management strategies in their operations.

Business support services in water utilities are undergoing DT at about the same pace as in other industries, especially similar ones like electric power. New information systems are enabling possibilities to serve customers better while at the same time implementing water conservation and other needed management programs.

Technical management methods using digital tools offer advantages for many utilities because they can leverage system data for purposes like maintenance management, system renewal, and water loss control. GIS has proved to be a powerful tool for such digital management because spatial and system data located in one place provide the needed data for multiple utility management processes. Also, utilities can integrate data management by locating their GIS systems centrally to leverage data and workforce capabilities jointly, as in responding to work orders, repairing breaks, and addressing other infrastructure issues. Utilities favor commercial packages to organize their data due to their lack of workforce capacity for research and development of data management systems.

GIS spatial data combined with system data provides the basis for asset management systems. They are promising, but the relatively slow adoption of asset management signals that new ways to implement it are needed. The method is data intensive and aligns well with DT, but this challenges utilities to do better in collecting and managing system data by integrated approaches across organizations. Team-building approaches will be needed to make improvements in many utilities.

In the operational world, SCADA is ubiquitous and can be paired with control strategies based on models, sensors, actuators, and management oversight to keep systems functional and effective. Cyber security will continue and even grow as an issue, especially for utilities with workforce and financial challenges. SCADA systems can be part of digital twins, which seem to offer advantages of enterprise data, decision support, and simulation models. Once capabilities are in place for upgraded versions of SCADA systems, distribution system optimization should be implemented more widely. Digital twins can integrate AMI devices and methods like water loss control through more granular pressure control. Even models to anticipate breaks can be integrated within them.

To underscore their importance and promise, GIS and SCADA are proven DT tools for water utilities, and other management systems and methods are evolving. Digital twin packages also promise to be an effective integrating method to help with workforce challenges while improving management outcomes. These technologies, together with new approaches to leverage data analytics, seem to point to the areas of future emphasis in DT of water utilities.

Researchers continue to integrate and synthesize data systems, but how rapidly and broadly utilities will adopt new measures is uncertain. The WRF has reports to help utilities with DT, but not all utilities have access and, even when they do, the research studies may need further work to be implementable. DT is complex and has many parts that make it difficult for utilities to implement on a comprehensive basis. Utilities will need help from their support community of regulators, consultants, vendors. and researchers to navigate the pathways that lie ahead.