My PhD investigates how AI can help us understand both ecological and social dimensions of large-scale infrastructure, with a particular focus on utility-scale solar energy (USSE). The environmental challenge is to balance rapid UK solar deployment with biodiversity protection and social legitimacy. My aim is to generate evidence that supports wildlife-friendly design of USSE and better incorporates the values of communities.
The first strand examines how bats and birds adapt to utility-scale photovoltaic (PV) developments. Using passive acoustic monitoring, I am gathering large audio datasets from solar farms and surrounding landscapes. AI-based classifiers let me process and analyse these data at scale, revealing species-specific responses and changes in community composition with distance from arrays that would be impractical to derive manually.
I am working with solar-farm asset managers and private landowners, feeding back site-level results to inform habitat management and engagement with environmental stewardship schemes. I’ve completed the first year of solar field data collection and begun processing the acoustic datasets; next steps include a further sampling season in 2026 at new sites, followed by spatial analyses. A highlight so far has been building relationships with asset managers and landowners. Despite divisive debates on solar–biodiversity, support for the research has been open and energising.
The second strand develops monitoring methods for hard-to-detect nocturnal species. I am creating AI-assisted computer-vision software to detect and track bats in night-vision video recorded within solar farms. This tool enables efficient monitoring of interactions with panels, helping assess whether maladaptive behaviour is a concern and, if so, how targeted interventions might mitigate it. I’m currently finalising the bat-detection and tracking tool and beginning to process the solar farm video data.
The third strand explores community values towards USSE. Using large language models for textual analysis, I will examine public consultation material to reveal recurring themes in consultation responses to nationally significant infrastructure. This is due to commence early 2026.