What if events were quantified, not lost?

In 2018 we learned a lesson: AI models are only as good as the data underneath them.

So we leaned into cleaning and labelling SCADA data. Hence the name SensorClean.

Today we have a machine learning event finder for SCADA water network monitoring data. 

These models find data signatures such as Pipe Failure Downstream, Sensor Offline and others.  

The figure below shows an example of SensorClean’s Event presentation - which forms part of our clients’ improved approach to managing non-revenue water. 

The figure shows 6 subplots all with diurnal profiles and identified events shown in different colours.

System 24 - Example of identified events that are found and quantified for evidence based decision-making

The value is in the ranking, which systems are costing the most, and the evidence for why.

Previous
Previous

Where’s your night flow baseline?