The role IoT and AI has in the digital transformation of water utilities

Water utilities that traditionally enjoyed a monopoly in many regions with a captive customer base are compelled to evolve due to several factors, including regulatory policy changes, impacts of climate change, and increasing consumer expectations.

The United Nations World Water Development Report published in mid-2019 estimates that 3.6 billion people live in areas that suffer from water scarcity at least 1 month each year and, by 2050, 6 billion people around the world will suffer from scarcity of clean water. To combat these challenges, water utilities are leveraging the Internet of Things (IoT) and their operations for efficient water management throughout the water cycle – from water resources to the taps, and efficient wastewater treatment to protect water sources. Water utilities are at a critical juncture, needing to upgrade existing infrastructure to improve resilience of critical operations to serve their consumers and protect water resources, while governments and regulatory policymakers are undertaking several short-term and long-term measures to improve regional water security and self-sufficiency.

Key challenges and the impact of the IoT

Non-revenue water is a significant issue for water utilities, occurring from leaking and broken pipes that tend to be caused by either outdated infrastructure or poor maintenance. According to a report published by water management services provider Veolia in 2017, the average efficiency rate of drinking water networks was 73.7%, resulting in a 26.3% loss of non-revenue water annually. Some utilities are tackling the challenge of non-revenue water by including leak reduction goals. For example, the city of Lille in France has a target of improving its water network performance by 6%, increasing from 79% to 85% by 2023 using IoT based water management solutions. Other regional governments are also directing utilities to upgrade their water distribution networks through regulatory policies. Some of the early notable initiatives from countries like Singapore, South Korea, Israel, and Malta, where regulatory policymakers have mandated the use of digital technologies to improve smart water grids and to reduce utilities’ water losses to less than 12%, are increasingly driving adoption of leak detection, and monitoring water pressure.

In South Korea, Gochang Waterworks implemented smart water meters in 24,000 households in Gochang County by the end of 2017. This smart metering project not only improved the accuracy of usage-based data for billing, but also resulted in reducing costs from leakages by 19%. An unintended outcome from the Gochang Waterworks’ smart meter deployments has been the utility’s ability to provide increased safety for its elderly customers by alerting relatives or triggering welfare checks in households without water usage. Similarly, Ofwat, the water services regulation authority in the United Kingdom, has mandated that water utilities reduce their water leakage by 15% by 2025, forcing water utilities to undertake initiatives to digitize their distribution infrastructure. In mid2019, South East Water announced trials in partnership with industry experts to develop and connect smart water meters, and placing acoustic sensors on underground mains pipelines using Vodafone’s Narrowband-IoT (NB-IoT) network to precisely detect and prevent leaks in their distribution system. South East Water is implementing Xylem’s Visenti for software analytics to manage and analyze sensor data installed on water systems, such as flow rate, level, volume totalizer, pressure, and water quality. To address the water utilities’ growing demand for IoT solutions, regional telcos are building their capabilities by partnering with IoT platform vendors with expertise in the water sector. In March 2020, Telefónica announced its partnership with Idrica to develop solutions for the digital transformation of the water utility’s operations. As the technology hardware and software ecosystem matures, water utilities are investing in IoT platforms based on a cost-benefit t ratio and capital investments that do not put a significant strain on their financial budgets. Typically, water utilities begin their digitalization journey by either automating their demand-side operations, such as water distribution networks, and/or some critical supply-side operations, such as monitoring their water resources that include dams, reservoirs, and water treatment plants. However, water utilities are increasingly implementing smart meters due to their various benefits, and to the relative maturity of the solution’s ecosystem. The costs of hardware and networks have also decreased with the commercial availability of dedicated public and private Low-Power Wide-Area (LPWA) networks. ABI Research estimates that water utilities’ smart meter installations worldwide will witness a 28% cumulative aggregate growth rate to reach nearly 400 million units in 2026.

Impact of AI in developing an integrated water resource management system

As water utilities deploy smart meters, sensors, and other IoT hardware, utilities will inevitably handle increasing amounts of data. Smart meter data, for example, are stored in Meter Data Management (MDM) platforms that are specifically designed to manage large amounts of data from millions of devices that are stored and processed according to specific requirements of the utilities. MDM software offers utilities a horizontal platform for integrating common data resources (smart meter data) across multiple applications, such as billing, asset management, and field service management. Such horizontal data management platforms that are owned and used by various business units within the utility, facilitate the smooth flow of business operations, thus increasing overall efficiency and reducing costs.

The recent COVID-19 pandemic has put significant strain on water utilities’ ability to continue delivering critical services to customers while operating with limited resources. As we come out of the pandemic, utilities should make their supply chain more resilient to future shocks by leveraging digital tools to optimize and automate many of their water management operations beyond smart metering and meter-to-cash applications. Utilities should develop a long-term holistic vision of an Integrated Water Resource Management System (IWRMS) that acts as a central system of record and a control system for all of their assets. Finally, water utilities should partner with technology service providers and system integrators to carefully evaluate innovations in Artificial Intelligence (AI) and Machine Learning (ML) technologies that help to efficiently process data from multiple sources in real time into actionable operational insights.

AI is here defined as intelligence demonstrated by machines with the attempt to simulate or replicate human behaviors and capabilities. In the past, sets of “if-then” rules (representations of individual concepts, stereotyped action sequences, and semantic networks) were some of the models adopted for AI systems. In recent years, ML has started to gain momentum. ML is a subset of AI that describes the process of creating, training, and executing a computer program with the explicit objectives of improving the analysis of a given task and obtaining measurable performance results.

As utilities evaluate various IoT platforms, they will also need to decide whether to invest in AI infrastructure in the cloud or at the edge. Nearly all AI training and inferences are performed in the cloud. Due to a cloud network’s scalability and flexibility, many organizations have chosen to rely on cloud computing, storage, and networking architecture. Some cloud AI applications include the integration with utilities’ Enterprise Resource Planning (ERP) or Customer Relationship Management (CRM) systems – such as billing loop systems, customer chatbots for customer service, inventory management, and field service management operations – that improve efficiencies and reduce manual interventions. As the AI and ML solution ecosystem matures, the execution of AI functions will move closer to the edge devices. This move will be driven by cheaper edge hardware, a lack of reliable and cost-effective connectivity options, enhanced data security, latency-sensitive businesses, and mission-critical applications. Edge AI will also enable utilities to create automated closed-loop systems that continuously monitor the health of critical infrastructure to enable efficient asset management through predictive maintenance and to build necessary redundancy in the system to reduce downtime and increase reliability. Using various remote sensing and imaging solutions in conjunction with geographic information system technologies, utilities can map and monitor their water resources. Using smart meter data provides accurate insights on end-customers’ water consumption and improves the accuracy of demand and supply-side forecasting. However, it is important to note that today, as the level of AI complexity increases for more accurate insights, there is a diminishing return in terms of cost to value.

In conclusion, the IoT, AI, and other emerging digital technologies have the potential to transform water utilities by improving day-today water management and addressing longerterm challenges of water security and resilience to natural disasters and climate change. The IoT and AI will also play an increasingly important role in planning and designing water microgrids for decentralized water and sanitization systems. As water utilities embrace the IoT, it is important to develop a roadmap for digitalization, while maintaining customer and business outcomes as key focal points.

Via Smart-Energy.com
By Adarsh Krishnan