A lightweight, autonomous water-quality sensor that draws power from the currents it monitors has taken out the global Sustainability prize at the 2025 James Dyson Award.
The device, called WaterSense, was created by Polish nanotechnology researcher Filip Budny, a PhD candidate at the Warsaw University of Technology. His idea: to replace manual, sporadic water-sampling in rivers and lakes with continuous, AI-assisted monitoring capable of early pollution alerts.
“More than 40% of the world’s rivers and lakes are seriously polluted, yet most are still monitored using outdated, manual methods,” Budny says. “I wanted to build something simple and self-sustaining that gives communities a live picture of their water quality.”
WaterSense is powered by hydrogeneration: a small generator inside its housing converts natural flowing water into electrical energy. The unit floats or anchors at adjustable depths and draws samples from three levels in the water column, enabling layered analysis of hidden pollutants. Instead of metal or plastic probes, the system uses recyclable paper sensors to monitor more than 20 indicators, including pH, dissolved oxygen, nitrates, chlorides and conductivity. Each day a fresh section of sensor paper advances automatically, while spent sensors remain sealed inside until the annual roll replacement.
Data is transmitted via mobile networks at set intervals – every minute, every 15 minutes or hourly depending on the setting –and delivered to an AI-powered dashboard which analyses patterns and forecasts pollution events up to 72 hours ahead. Results are made publicly available, giving local authorities and communities actionable insights in near-real time.
From data to action
According to Budny, the goal is early detection, not just measurement. “When something starts to change – such as a rise in nitrates or a drop in dissolved oxygen – the system can alert councils or local authorities immediately. That gives them time to locate the source before damage spreads.” He adds that community access to the data builds shared responsibility across stakeholders. Initial real-world trials are underway in Poland at 20 sites in collaboration with water companies and government agencies; the next step is a Europe-wide network with ambitions for a global rollout by the end of the decade.
Why New Zealand should take notice
The relevance of continuous monitoring becomes especially clear in the context of New Zealand’s freshwater context. National reporting shows more than 1,700 rivers and streams are monitored across the country, but many sites record data only monthly or annually. Land Air Water Aotearoa (LAWA) reports that almost two-thirds of those monitored waterways are graded C or D in terms of ecological health.
The fact is that despite decades of management effort, agricultural intensification and land-use pressures continue to degrade water quality in some catchments. A recent study by the National Institute of Water & Atmospheric Research (NIWA) noted that although wastewater-treatment upgrades have improved results in some places, diffuse pollution from farming and forestry remains difficult to control.
With regulation tightening – such as the National Policy Statement for Freshwater Management 2020, which sets ‘bottom lines’ for nutrient and microbial contaminants – the need for higher-frequency, automated monitoring is rising. Traditional manual sampling cannot reliably detect rapid changes in water condition. One analysis found that to detect changes in E coli within five years rather than 20, sampling frequency would need to increase five-fold.
WaterSense offers the kind of continuous, depth-resolved, and AI-analysed monitoring New Zealand needs. With the ability to deploy across catchments, lakes and estuaries and feed data into national-level systems, it could strengthen regional council-led monitoring and support more responsive water-management strategy.
Engineering the difference
WaterSense’s innovation is not just in data but in mechanical and systems design. By generating its own power and automatically replacing sensors, it addresses two major barriers to sustainable deployment: energy supply and maintenance logistics. The use of paper-based sensors lowers cost and environmental impact, enabling wider deployment in remote or under-monitored catchments.
Professor Tamara Tokarczyk of Poland’s Institute of Meteorology and Water Management describes the system as a “game-changer”. “WaterSense is the first solution to deliver systematic, real-time qualitative monitoring. The high-resolution data it collects will significantly improve forecasting models and benefit both society and ecological systems.”
In engineering terms, the unit’s real-time data feed enables predictive maintenance of infrastructure and faster response to pollution events. For New Zealand water engineers and asset managers overseeing rivers, irrigation networks and storm-water systems, that means fewer blind spots and fewer reactive interventions.
Recognising and scaling impact
The award-winning status of WaterSense came with a prize of US$66,100 (about NZ $106,000) and wide media exposure. Budny said the prize gives the project momentum to refine prototypes and seek investment for full production: “The James Dyson Award confirms that environmental innovation has a global voice,” he says. “The prize allows us to refine prototypes, gather more data and move toward production. We are now working with partners and investors to scale WaterSense into a Europe-wide network by 2026.”
Future potential
Although the current focus is on rivers and lakes, Budny sees broader applications: coastal zones, reservoirs and industrial outflows. The sensor roll could be adapted to track metals, microplastics or emerging contaminants and the AI platform could integrate catchment-scale data streams. The ambition is a network that enables water systems to self-monitor and warn stakeholders before pollution becomes a crisis.
For New Zealand’s engineering community, WaterSense illustrates how low-cost materials, self-powering designs and smart data can converge to solve real-world ecological challenges. Developing this kind of technology in a national context means operability across diverse terrains, integration with regional council monitoring systems and alignment with mātauranga Māori perspectives on wai. Given the scale of the freshwater challenge, systems that deliver continuous, high-resolution data could become infrastructure as important as pipes and pumps.
Budny concludes: “Clean water is essential to life, but we cannot protect what we cannot measure. WaterSense is about giving every river a voice.”