AI-Powered Bird Monitoring at Offshore Wind Farms Shows Promise

The trial of AI-powered vision systems to monitor bird movement at offshore wind farm, conducted by Spoor AI, Vattenfall, and the British Trust of Ornithology, has proved promising.

The trial was conducted at Vattenfall’s Aberdeen Bay Offshore Wind Farm.

The study, a combination of theoretical assessments and field trials, evaluated how two different camera systems and Spoor AI’s video processing software detected birds and their movements around wind turbines.

The trial compared two different camera systems - standard mono-vision and stereo-vision. Compared to the mono-vision system, which estimates bird distances based on assumed bird sizes, the stereo-vision system delivered more accurate distance measurements as it used two cameras to triangulate distances, similar to human depth perception.

This means that the size of the bird is more accurately assessed, and therefore its movements recorded more precisely.

This level of detail provided by the stereo-vision system could allow wind farm operators to document ‘micro-avoidance behavior’. Micro-avoidance behavior of birds at wind farms refers to the last-minute flight adjustments birds make to avoid colliding with the rotor blades of turbines.

This behavior is part of a broader spectrum of avoidance strategies that birds employ when navigating through or near wind farms.

According to Vattenfall,