Utilities · Anomaly detection
Sensor data cleaning for water networks
Network telemetry cleaned before anyone acts on it
The problem
Water networks report pressure and flow through sensors that drop out, drift and spike. Analysis of losses and leaks is only as good as that data, and cleaning it by hand does not scale.
What we built
We built a platform that ingests network telemetry, detects anomalies such as gaps, drift and spikes, and cleans or flags the readings before they are used to analyze network losses.
What changed
- Sensor readings cleaned and anomalies flagged before analysis
- Ran on utility data from 2021 to 2023
- Wound down at the client's request, with the code kept
Have a problem in this shape?
Tell us what you are trying to win. We will map your version of this build in one call.