Elvia supplies electricity to 2 million people. This requires 65,000 km of cable and 440,000 poles. The poles are subject to wear and tear, so to prevent power outages, they are regularly inspected by helicopters that take pictures and map damage. Computas was tasked with creating an image platform where the images are automatically analyzed to detect faults in the power grid.

Previously, the images were manually analyzed by engineers and the files were handed over to Elvia on hard drives. In the “Louvre” project, Computas collected all images in a single archive. A machine learning solution was then added that can detect faults on the poles just seconds after the images are taken.

“From a purely technical standpoint, this is a very exciting project. We have created a plugin framework where the systems that should process the images are plugged in. The framework has a rule engine that allows different systems to register which images they are interested in,” says Sigrid Andersen, developer at Computas.

This solution enables Elvia’s engineers to use their time more efficiently and ensures that millions of others can take it for granted that there is light and heat every day, all the time.

Louvre is just one of many systems that Computas has contributed to at Elvia. Computas has also developed Elvia’s website, power outage map, automatic detection of faults in the metering network, and a platform that makes it easy to do things right when it comes to development, deployment, and operation.

“The joy of creation is the foundation of everything we do, and here we have combined knowledge and creativity to create something entirely new. I am passionate about making things better and easier for our customers. It is incredibly inspiring to be a part of – and I look forward to what comes next!” concludes Jonas Bjelkerud, Tech Lead on Team Glue, which develops the Louvre system, and site manager for Computas at Elvia.