Startup Touts New Technology to Predict Outages

Image courtesy of Stuart Rankin under Attribution-NonCommercial 2.0 Generic Deed, resized to 700 x 391 pixels.

Outage restoration and recovery gets all the proverbial press, but what if utilities could predict outages and address them before they actually occur?  That is the mission of Toumetis, a startup company that combines predictive analytics with machine learning to help utilities predict outages and enhance grid reliability.

Utilizing Advanced Software to Predict Outages

The company’s software platform, called Cascadence, analyzes power quality to predict power outages before they happen.  Whenever the software detects a power quality issue, it identifies the root cause and location of the anomaly.  This not only prevents a percentage of outages from occurring in the first place, but it can also reduce the duration of both planned and unpanned outage events.

This obviously provides huge benefits in terms of reliability and safety metrics, customer satisfaction ratings, and operational costs.  That said, it’s not exactly a plug-and-play solution, as the company must first work with the deploying utility to custom-build the data pipeline.  But once the pipeline is built, the technology can run seamlessly – live data is fed into the machine learning algorithms, and anomalies are automatically identified in real time by equipment type and location based on the previously-built requirements.

So far, Toumetis has deployed this solution for San Diego Electric & Gas to great effect – dozens of outages have been predicted and prevented during the pilot.  The company has also partnered to develop iPredict, which is an artificial intelligence (AI) platform that can predict asset failures weeks before they actually occur.  It is also developing a similar platform for underground cable monitoring.

I absolutely love this idea!  Simply put, if more outages can be prevented via predictive analytics, the process of utility emergency preparedness would become infinitely more effective and efficient.

In the final analysis, the ability to predict outages, thereby allowing them to be prevented, is invaluable when it comes to maximizing grid reliability, customer satisfaction, and safety, and reducing operational costs.