The future of water sensing technology
Our award-winning technology utilises deep learning on-device (tinyML) enabled by state-of-the-art ultra-low-power neural decision processors and sensors to drive superior flow detection performance on battery power alone. It’s delivering a step change in water monitoring.
The non-invasive battery-powered sensing technology can be deployed across the built environment at scale - paving the way for more efficient water monitoring and increased sustainability.
Deep learning model for flow detection
Refined over six years of research and development, our technology uses deep learning models to recognise the unique and complex patterns of high-fidelity multivariate sensor data when water flows through a pipe.
Deep learning aims to mimic the human brain, with models trained to recognise complex patterns and provide relevant and reliable insights and predictions.
Our models are pre-trained on an ever-growing multi-terabyte pool of reference flow and multivariate sensor data. They are highly sensitive and can reliably recognise water flow down to a single-second resolution.
Unrivalled performance, ultra-low power
Our deep learning technology accurately identifies and reports flow events at ultra-low power. Devices can operate for years on a single battery.
Harnessing state-of-the-art neural decision processors, our deep learning models are executed directly on the sensor device (tinyML).
This innovative approach eliminates the need to transmit large volumes of sensor data for processing in the cloud, saving energy while increasing speed and security.
Installation freedom at scale
Battery-powered devices can be installed in minutes and operate where needed without mains power or a wifi connection. The small devices fit to existing pipework at the point of use without cutting through pipes.
Data is processed on-device and transmitted via LoRaWAn, a secure long-range radio transmission protocol designed specifically for IoT applications. The system is particularly suited to dense urban environments where data from hundreds of devices can be delivered directly to a customer dashboard.
The technology we have created has been used to develop our first product, Cortense. Cortense combines our ultra-low-power sensing technology with machine learning to create the industry’s first device capable of monitoring both water flow and temperature reliably and efficiently. It’s the first device of its type available on the market and could transform Legionella risk management and cut water waste.
