DHI biologists film bats and birds at night to train MUSE AI system for Ecowende wind farm
Friday 6 December 2024
A team of DHI biologists recently collected field data on bats’ and birds’ movements at night. The data will be used to train DHI’s MUSE artificial intelligence (AI) solution to automatically discriminate between bats and birds at night. MUSE will be deployed at our offshore wind farm to monitor and protect bats.
DHI’s MUSE system integrates any combination of radars, digital cameras and acoustic sensors to detect, monitor and protect birds and bats while maximising uptime of the wind turbines. MUSE AI for bird species recognition during daytime was launched a few years ago. The next step is to enable the MUSE AI algorithm to recognise and discriminate between bats and small birds during the night, using the new recordings as training data.
Minimise impact on vulnerable species
Most bats are of species that are considered vulnerable and are protected by the EU Habitat Directive, and European offshore windfarms must be able to document how they will minimise the negative impact on bats when in operation. As some bat species migrate between the Netherlands and the UK, they risk encountering and potentially colliding with turbines in an offshore windfarm. In fact, though there is a lot of knowledge about bats on land, there is still much to learn about bats at sea. As such, we want to gain as much knowledge as possible about these species to better understand how one should mitigate the impact on bats at sea. Innovations from DHI, such as MUSE coupled with thermal cameras, will support our monitoring and research programme.
Bats are an important part of the ecosystem
Some species of bats are predators of common nocturnal insects and regulate these populations. Across Europe, there can be found around 45 species of bats, occupying a wide range of habitats. On land, they each face a variety of pressures such as landscape change, agricultural intensification and habitat fragmentation.
Filming in Denmark’s bat hotspots
The field team of DHI biologists borrowed a cooled thermal camera, and subcontractor Circle Consult helped the team set up the advanced equipment for several nights in one of Denmark’s bat hotspots in the region Lolland-Falster. DHI MUSE AI experts will use the recordings to train the AI to automatically distinguish between bats and small birds in low light. This will eventually help to reduce bat collisions. The DHI MUSE scientific team can create more detailed analyses and potentially discover unseen behavioural patterns of bats at night. Ultimately helping to protect this vulnerable species group.
