Water utilities worldwide are at a turning point, navigating digital transformation in a persistently disruptive technological environment.
In a recent webinar hosted by Paul Fleming, experts Gigi Karmous-Edwards, Dragan Savic, and Saša Tomić explored the impacts of emerging technologies like AI, digital twins, and robotics on the water sector. During an exciting discussion, the panel delved into opportunities and challenges in integrating these technologies to optimize water management.
Generative AI: Opportunities in the water sector
Generative AI, a technology reshaping industries globally, is also poised to revolutionize the water sector. Unlike traditional AI, Generative AI leverages natural language processing and large language models (LLMs) to generate new content, such as text, images, and videos. Gigi, a self-professed AI evangelist in the water sector, emphasized how this technology can help utilities achieve their digital transformation goals.
She believes that Generative AI is not poised to take jobs but to transform them. She said, “It’s not a matter of how generative AI will start automating everything. It’s more, how can we use its capabilities very slowly and very surely?”
She also mentioned Retrieval Augmented Generation (RAG) — a technique that combines LLMs with external knowledge bases and data to improve their outputs — and its potential to help those in the water sector create large, verticalized language models tailored to specific needs.
However, she acknowledged that adopting Generative AI in the water sector comes with challenges. Many utilities are cautious about data privacy and cybersecurity, while many employees fear job displacement. She suggested starting with small-scale experiments, like deploying Optical Character Recognition (OCR) for digitizing paper-based records. Using AI in this way shows its real benefits and helps build confidence among teams.
Reflecting on over 30 years of experience with AI, Dragan warned against getting caught up in its hype. “Expecting too much, too soon, can backfire,” he warned. He advocated focusing on areas where it can be directly applied. The panel identified an example: KWR, the organization Dragan works for, has amassed more than 10,000 research reports over the past 40 years in the water sector. AI could analyze this huge archive and improve utilities’ knowledge management processes.
The continuous evolution of digital twins
Saša traced the evolution of digital twins (virtual models that integrate data and simulation for decision-making), explaining their evolution.
He outlined how the introduction of SCADA and GIS systems in the 1980s and 1990s enabled water utilities to integrate data for basic decision-making, like turning on a pump when a tank was low. This transition from raw data to meaningful information was a key step in digital transformation.
Next came the shift from information to knowledge, where systems like GIS dashboards synthesized data with real-world experience to aid decision-making. Over time, these systems evolved to offer intelligence, combining real-time data and predictive models to anticipate changes and make smarter decisions.
Saša had a simple caveat: “We must keep humans in the loop for critical decisions. Let machines handle the repetitive tasks; leave creativity to humans. AI, in fact, will allow us to be more creative.”
Today’s digital twins are more powerful and accessible, thanks to the integration of AI and large-scale data systems. Gigi highlighted how generative AI can improve these systems by democratizing access to insights. For example, operators can use natural language queries to ask specific questions and receive clear, actionable answers. This accessibility simplifies complex technology for those without expertise in modeling or dashboards, allowing more stakeholders to participate in decision-making.
Gigi and Saša also discussed including unstructured data (videos, images, and operator fieldworker notes) in the knowledge-gathering aspect of digital twins. This development bridges gaps in knowledge and creates a holistic view of water systems that enables utilities to make informed, proactive decisions.
Pushing boundaries: The integration of robotics and AI
Dragan turned the focus to robotics as a transformative tool for managing underground water infrastructure. Traditionally, accessing and assessing buried pipes has been labor-intensive and challenging. Robotics, combined with AI, offers new possibilities for addressing underground infrastructure challenges, from mapping pipe conditions to detecting sediment build-ups and anomalies.
“Imagine autonomous robots capable of identifying and removing fatbergs in sewers before they cause blockages,” said Dragan, highlighting how this kind of innovation could save utilities millions in maintenance costs.
Gigi expanded on this vision by exploring the integration of robotics and computer vision. She described a scenario where robots continuously photograph and analyze wastewater sludge, optimizing treatment processes in real-time. This integration could help utilities achieve constant, optimal water management, improving efficiency and reducing costs. “This fusion,” she explained, “moves us closer to constant, optimal water management.”
The centrality of data to digital transformation
All panelists agreed that data is the foundation of digital transformation. However, the water sector must overcome hurdles like siloed systems, poor data quality, and lack of standardization.
“Clean, structured, and integrated data is as valuable as any pump or pipe,” said Saša, stressing the importance of treating data as a critical asset.
Gigi agreed, pointing to the need for standardizing asset metadata across utilities. By adopting universal frameworks, utilities could share and integrate data more effectively, creating opportunities for collaboration and innovation. She cited initiatives like British Water and WOSA’s BIM classification systems as promising examples of standardization efforts.
Looking ahead
The combination of AI, digital twins, and robotics offers huge potential for the water sector, but as the panelists pointed out, it’s important to plan carefully, manage data well, and find the right balance between automation and human input.
Explore more insights — listen to the webinar in full!
If you want to dive deeper into this discussion, give the full webinar a listen and explore even more insights from the panelists.
