Advanced Machine Learning Analytics
StormHarvester Blockage Predictor detects blockages across sewer networks before they cause a flood or pollution event. It provides a platform to monitor all sewers in a network using existing level and flow sensors.
This tool means Operations room staff will no longer be overwhelmed by the high numbers of alarms during wet weather events.
Blockage Predictor silences alarms for consented events and only provides alarms/alerts for unconsented events so maintenance crews can quickly be directed to incidents of greatest impact. This allows Operations staff to move from reactive control to proactive control.
How Blockage Predictor Works.
Using StormHarvester machine learning prediction software, the tool learns normal behaviour of a site. From this, we create safe operating windows for each sewer level monitor, across the sewer network.
If the upper or lower threshold of this operating window is breached, real-time alerts for a possible blockage/anomaly are delivered.
Specific alerts enable rapid response and repair – reducing network outages and their potential environmental impact.