Protecting endangered birds with computer vision

Regulations in many countries require wind farm operators an increased tracking and documentation of bird behavior around wind farms. Previously birds have been tracked by onsite human personnel, and when rare species or large flocks of birds approached, entire offshore wind farms needed to be shut down. MUSE, the multi-sensor bird detection application by DHI, offers developers and operators frequent, high-quality observations based on an integrated setup of radars and cameras. Real-time bird behavior is automatically recorded and displayed online, ensuring regulatory compliance and eliminating the need for onsite staff. Together with DHI, Silo AI built an AI-based species recognition solution that automates the process further and enables speed reduction of the individual turbines the birds are approaching, eliminating the need for  shutting down the entire farm, resulting in more energy and cost-efficient operations.

AI-based species recognition enables the shutdown of individual wind turbines

DHI and Silo AI built AI functionality for DHI's multi-sensor bird detection application. The computer vision solution enables hierarchical classification of the birds enabling highly accurate recognition of endangered bird species and large flocks of birds. The algorithms calculate the birds' flight routes, prompting an automatic shutdown that can be limited to single turbines. Limiting the shutdowns to only singular turbines helps offshore wind farms optimize their energy consumption, reducing costs and making operations significantly more environmentally friendly, allowing more efficient green energy production.

‍