Why The Water Sector's AI Ambitions Depend On Better Connectivity
By Eric Verheylewegen

Recent research underlines just how quickly AI is becoming central to modern utility operations, with Singapore, China, and Japan already using smart technologies to detect leaks and protect infrastructure. Meanwhile, the UK is said to be perhaps a decade behind: with more work required to prevent treated water leaking into the surrounding environment.
For AI to deliver real operational value, it needs a constant flow of reliable operational data. AI systems are relentlessly data-hungry, and the more data, the better. Yet, accessing this data remains a major challenge in the utilities sector, with remote reservoirs, wastewater treatment works, and sprawling infrastructure often located a long way from traditional cellular networks.
This is where emerging connectivity applications using Narrowband Non-Terrestrial Networks (NB-NTN) could be a key part of the solution. NB-NTN, utilizing industry-wide 3GPP network standards, can allow IoT devices to connect to both satellite and cellular networks through a single SIM and chipset module, without the need for traditional satellite terminals. This can make reliable connectivity much more cost-effective to deploy and easier to integrate across vast, hard-to-reach operations.
That, in turn, could help water companies collect the data AI systems require; supporting faster, more accurate decisions, and implementing new leak detection models.
Underlining the rising demand for better connectivity to transform the sector’s AI ambitions into operational reality, recent research from Viasat reveals that 63% of utility leaders plan to adopt this solution within the next 12 months. While full deployment will take time — 91% expect that to happen over the next couple of years — the direction of travel is clear.
The Dual Challenge: Aging Infrastructure And Connectivity Gaps
The challenge for utilities companies and their infrastructure is twofold: it is aging and remote. This combination results in hard-to-detect leaks, bursts, and inefficiencies which are difficult to manage proactively, leading to reduced supply, costly disruption, and even pollution in some cases.
Without real-time monitoring across these vast networks, it's tough to pinpoint where and why problems occur. This has understandably led to a reactive approach, with companies playing catch-up to fix problems, rather than preventing them. The overhaul of UK water regulation earlier this year has brought renewed scrutiny and tougher expectations around water quality and leakage reduction, with significant financial penalties for failure.
AI and big data analytics offer a path to more proactive, preventative management, but it is intensely data hungry. With a constant stream of data, AI platforms can detect leaks reliably, predict when and where pipes might fail, and identify the weakest sections of the network for preventive maintenance. By moving from a reactive to a proactive model, firms can reduce waste, avoid costly environmental fines, and build a more resilient and cost-efficient network.
But AI is only as effective as the data you can gather. That data is then worthless if you cannot reliably connect all your assets and get a clear picture of what is happening on the network. By nature, many parts of the UK’s water network are in remote areas where terrestrial connectivity is limited or non-existent, creating a significant barrier to deploying these smart solutions at scale.
Bridging The Connectivity Divide With NB-NTN
NB-NTN is a potentially game-changing technology that enables IoT devices to connect to both cellular and satellite networks through a single, integrated module, without requiring a traditional satellite terminal. This reduces both complexity and cost of deploying IoT at scale and alters the cost/benefit analysis for organizations. For the UK's extensive water infrastructure, a constant, reliable flow of data from nearly any asset, anywhere, could now be in sight.
For the water sector, this means a more complete picture of assets across the network, wherever they are located.
Unlocking The Full Potential Of AI With Real-World Applications
By providing a reliable stream of data and reducing complexity through standards-based innovation, NB-NTN acts as the channel through which AI systems can access a vast reservoir of data and truly transform water management.
Our research found that water infrastructure monitoring was ranked as a top transformative use case by 43% of utility leaders. In practice, that means sensors can track flow and water quality in real-time, flagging leaks or contamination as soon as they appear. Beyond monitoring, this connectivity also enables proactive control, allowing operators to remotely manage pumps and valves without costly and time-consuming site visits, which is crucial for managing resources during droughts.
The benefits extend to health and safety, another key priority for the sector. With 41% of leaders identifying 'people tracking' as a key use case, NB-NTN-connected devices can also ensure lone workers in remote locations are safe and can call for help in an emergency, regardless of cellular coverage.
This reliable, ubiquitous data stream is what makes the AI-powered predictive maintenance and forecasting a widespread reality.
Conclusion
The water industry is at an inflection point. The case for AI is increasingly clear, but so too is the need for the infrastructure that allows it to work properly. Without reliable remote connectivity, the sector will struggle to move from isolated pilots to widespread deployment.
NB-NTN provides a crucial piece of this puzzle, making satellite connectivity more accessible, cost effective, and scalable. It gives water companies a practical route to extend smart monitoring across their networks, enabling a reliable, ubiquitous data stream that empowers AI-powered predictive maintenance and forecasting - helping to build a more resilient, efficient, and sustainable water system.
Eric Verheylewegen is VP of Strategic Initiatives Enterprise and Land Mobile at Viasat.