Australian ecology specialists are using thermal imaging technology combined with artificial intelligence to unlock critical data about microbat flight behaviour at wind farm sites.
The breakthrough could reshape how the nation’s wind energy industry manages its impact on protected bat species.
Biosis, one of Australia’s leading specialist ecology companies, will present their latest research at the Clean Energy Council’s Wind Industry Forum on May 26.
The trial, already underway at four proposed wind farm sites in New South Wales, is delivering the kind of species-specific flight data that ecologists have long needed but been unable to obtain through conventional means.
Wind turbines pose a known and well-documented risk to bat populations across the globe.
In Europe and parts of North America, regulators have responded with firm requirements around seasonal turbine curtailment during peak bat activity.
In some cases, smart sensors capable of halting a turbine blade in real time when a bat approaches are also mandated.
Australia has lagged behind, not from lack of concern, but from a lack of data.
Australia is home to more than 60 microbat species, many carrying protected status.
As the nation’s renewable energy sector expands, pressure on wind farm developers and their ecological advisers to quantify collision risk has grown steadily.
Until now, that quantification has been elusive.
Traditional survey methods rely on audio detectors to capture the echolocation calls of bats near a site.
These detectors are useful for identifying threatened species, but they offer no visual evidence and no reliable indication of flight height, the single most important variable when assessing whether a bat is flying within the rotor sweep zone of a turbine.
They also cannot distinguish whether a single bat or a cluster is responsible for the calls being recorded.
The Biosis approach addresses these limitations by combining two technologies that, until recently, had not been deployed together in an Australian wind energy context.
Thermal imaging cameras capture the heat signatures of bats as they navigate the night sky, making their flight paths visible for the first time.
These visual records are then synchronised with ultrasonic audio recordings, allowing ecologists to match individual flight events to specific species through their echolocation signatures.
The result is a dataset that reveals not just which species are present, but where they fly, at what heights, and during which activity periods — including peak feeding times.
Biosis Senior Zoologist in Bat Ecology, Felicity Williams, has spent years conducting bat utilisation surveys at wind farm sites and has developed deep expertise in bat call identification.
She is leading the new data capture programme and has described it as transforming the way her team understands microbat behaviour in the wind farm environment.
To make this approach practical at scale, Biosis has partnered with Wildlife Imaging Systems (WIS), a US-based technology company that has developed machine learning software specifically designed to process thermal video feeds and extract meaningful biological data.
The WIS software takes dual thermal video streams, converts them into detections and tracks, synchronises the feeds, and performs the camera-to-camera calibration necessary to estimate the three-dimensional location of each bat or bird in the frame.
The system uses pattern recognition to distinguish bat tracks from those of birds, allowing Williams and her team to focus their analysis on the relevant subjects.
A key feature of the technology is its patent-pending auto-calibration capability, which allows 3D positioning to be estimated using only the detections captured during the survey itself, eliminating the need for expensive synchronised hardware or labour-intensive field calibration processes.
The cameras used are field-ready security models that can be deployed remotely, powered by solar panels and batteries, making them well-suited to the remote locations typical of Australian wind farm developments.
For Wildlife Imaging Systems, the Biosis trial represents an unusually valuable opportunity.
Most of the company’s work takes place at operational wind farms, where turbines are already installed, and the primary objective is to reduce fatalities in real time.
Working with Biosis on pre-construction sites means the data collected could directly influence turbine placement, rotor height specifications, and the design of bat and bird management plans before a single blade is erected.
The implications extend well beyond Australia. Flight height measurements gathered in 3D across Australian microbat species would contribute to a global body of knowledge about pre-construction collision risk assessment, information currently in short supply for most of the world’s bat species.
Biosis expects that twelve months of data collection will yield the collision risk analysis needed to give their clients and the broader wind energy sector a robust, evidence-based foundation for future project planning.
