Case Studies
We believe using AI, tech, and data for good means supporting those working directly to protect our environment - ecologists, land managers, to national agencies and more. Their boots-on-the-ground work is what truly matters!
Tools like THEIA reduce time spent at the desk and help accelerate vital restoration efforts. Through ongoing R&D, we’re continually growing our expertise in technology and data science - supporting everything from animal censuses and habitat monitoring with satellite data, to fire risk modelling, global weather data pipelines, and forecasting water temperature shifts that affect species like wild salmon and freshwater pearl mussels.
All our projects aren’t just technical achievements — they’re part of a wider mission to accelerate climate action & safeguard fragile ecosystems.
Puffin Monitoring
Nature deserves the best possible data, delivered quickly, accessibly and at scale. Click on the Puffin image left to watch our video into how we're building the future of conservation in Scotland.
Using our THEIA Monitor 3D feature, we collaborated with the National Trust for Scotland and NatureScot to create a virtual fieldwork platform that supports ecologists in monitoring puffin populations across UK colonies.
As part of the Scottish Government’s CivTech Challenge 10.3, we partnered with the University of Edinburgh’s School of Geosciences and the Airborne Research Innovation Facility to improve on existing monitoring techniques through advanced technology. High-resolution aerial imagery was captured over puffin colonies, and a 3D ‘digital twin’ of the landscape was generated. This virtual model allows ecologists to identify and tag puffin burrows as Apparently Occupied (AOB) - the standard metric used by ornithologists - or unoccupied.
To further support the process, we developed an AI algorithm capable of automatically detecting burrow locations, streamlining the counting process. This innovative platform also has the potential to be adapted for monitoring other seabird species.
Human Wildlife Conflict
We worked with Omanos Analytics and the Peace Parks Foundation to develop a satellite-powered data portal for monitoring human-wildlife conflict (HWC) in Mozambique’s Limpopo and Banhine National Parks.
Focusing on elephant populations, the project used high-resolution satellite imagery and machine learning to:
- Map human activity and elephant distribution
- Identify conflict hotspots and seasonal risk patterns
- Detect small-scale croplands often missed in standard maps
- Model seasonal elephant movements using environmental data
Thanks to THEIA Monitor, we provided a powerful, landscape-scale tool to support smarter, data-driven conservation and land management across the region.
Animal Census
Animal population data is essential for conservation, land use, and environmental protection. We deliver scalable, tech-driven solutions that replace traditional, costly, and carbon-intensive survey methods.
In Scotland, deer numbers are critical to managing natural assets. Working with a government-backed scheme, we used AI and 12.5cm resolution aerial imagery to detect and count free-roaming deer, offering a faster, lower-carbon alternative to helicopter surveys. [First image on the left.]
In Africa, satellite data enables elephant monitoring in remote or unstable regions. We map populations and compare them with historic ground data to assess the impacts of poaching, habitat loss, and human-wildlife-conflict.
Agricultural animals like cattle influence water quality and other elements. In collaboration with a Scottish environmental agency, we applied transfer learning to deliver a high-accuracy cattle detection model - supporting improved environmental oversight at scale.
These projects show how AI and remote sensing can transform how we understand and protect animal populations globally.
Watercourse Climate Resilience
With a warming climate, our freshwater habitats are facing increasing pressure due to rising temperatures, posing a significant risk to vital species such as Atlantic salmon and freshwater pearl mussel. EOLAS has taken action by developing a mapping and reporting service focused on watercourse resilience, helping organisations understand the effects of global temperatures on crucial local watercourses and the benefits of investing in mitigation strategies, such as riverside woodland schemes.
A robust machine learning framework designed to estimate water temperature across the Scottish river network is to be established, whereby the estimation will be applicable to any spatial point within the network, irrespective of the time of year. To achieve this, EOLAS is integrating optical satellite data, high-resolution digital elevation data, and climatic data (including rainfall and air temperature).
Canopy Detection
Our data scientists harness the power of AI to detect and classify woodland and scrub from aerial imagery. This automated process enables accurate, large-scale mapping of habitats, allowing for detailed analysis of change over time.
By observing how woodland areas evolve, we support better decision-making for conservation, planning, and land management. Whether tracking regrowth or assessing human impact, our AI-driven tools turn complex imagery into actionable insight - quickly, reliably, and at scale. It's smart tech with a green heart, helping protect and understand our natural landscapes like never before.
