Empowering rangers in South Africa with real-time thermal poacher detection to protect threatened species.
About the Project
In collaboration with the South African non-profit Strategic Protection of Threatened Species (SPOTS), Team Epoch is developing a critical AI tool to assist rangers on the front lines of wildlife conservation. During this 10-week sprint, the team will develop and deploy a real-time detection model that integrates with footage collected on a fixed-wing drone, reducing the workload for rangers and enabling a scale-up of patrols.
Organization
Strategic Protection of Threatened Species (SPOTS) is a South African NPO dedicated to the operational protection of wildlife. SPOTS works directly with front-line anti-poaching units, focusing on providing actionable intelligence and operational support to game reserves. They focus on initiatives designed to bridge the gap between cutting-edge technology and on-the-ground conservation needs, giving engineers and developers the opportunity to deploy their skills where they matter most.
Relevance
Poaching remains a critical threat to biodiversity across Africa, often driven by highly organized crime syndicates. Rangers tasked with protecting the vast reserves operate under extreme conditions, primarily at night. While traditional methods rely heavily on military-style training and manual patrols, there is a growing movement to integrate technology that keeps both animals and humans safe.
By utilizing drones, rangers can cover a massive operational radius. However, manually reviewing hours of thermal video in real-time is incredibly taxing and prone to human error. Team Epoch’s software bridges this gap by automatically flagging unauthorized activity in the video feed. This allows ground crews to respond before an incident occurs, creating a powerful deterrent and ensuring the ongoing protection of endangered wildlife.
Technical Details
The core technical challenge involves building a robust object detection model capable of identifying unauthorized activity using infrared video footage. The AI must perform reliably at night and remain effective across a range of operational conditions, such as varying altitudes, camera setups, and seasonal changes.
Because the system will be deployed in remote African reserves with limited to no internet connectivity, the model cannot rely on cloud computing. Instead, it will run locally on a mobile ground station, which places constraints on its size and complexity. Furthermore, because manual data labeling is a significant bottleneck, the team is developing auxiliary models to facilitate continuous improvement after handover. Finally, the software will be packaged into a streamlined, intuitive interface developed in close collaboration with local rangers, ensuring they can use the tool seamlessly in high-stress field conditions. At the end of the project, the team will test both the models and the interface on-site during an extensive in-person handover, to make sure the system makes a meaningful improvement to daily operations.
UN Sustainable Development Goals – 15 LIFE ON LAND
This project directly contributes to SDG 15: Life on Land, which aims to halt biodiversity loss and protect threatened terrestrial ecosystems. By equipping anti-poaching units with advanced AI technology, Team Epoch actively combats the illegal wildlife trade. The preventative nature of this software directly protects endangered species in vulnerable reserves, helping to preserve Africa’s vital biodiversity for future generations.
