Terra Drone has developed a power distribution inspection solution that combines beyond visual line of sight (BVLOS) UAV flights with artificial intelligence algorithms that is accurate up to 92.5 percent, the company said.

The algorithm was developed via identifying anomalies in a training data set of approximately 1,500 images. Terra Drone said the algorithm need only be completed once for a particular type of asset and can be implemented easily at different locations on similar assets. The system identifies anomalies such as rust on bolts and loosening and missing tower parts and then generates a report that highlights the areas that require action.

Clients can define security protocols and store data either in the cloud or on-premise. Terra Drone said the system provides for significant cost savings and risk reduction as opposed to traditional inspection methods of powerlines, towers, and substations such as low-flying helicopters or on-the-ground patrols.

The company said it developed its UAV/AI solution for the power industry after conducting more than 46,500 nm worth of BVLOS powerline inspection flights worldwide.