Early forest fire detection using drones
Forest fires pose an increasing threat to the environment, the population, and the economy, especially during the dry summer months. The project investigates how commercially available drones equipped with thermal imaging cameras and artificial intelligence (AI) can be used to detect potential fires early and precisely locate them. The goal is to develop a cost-effective and reliable alternative to existing monitoring methods.
The feasibility study represents the first phase of a three-stage project (feasibility study – prototype development – pilot project). During this phase, drone models, thermal imaging technologies, and suitable software development kits (SDKs) were evaluated. In addition, technical and organizational challenges such as flight planning, battery capacity, image resolution, and weather dependencies were analyzed, and solutions were developed.
A central element of the project is the automated processing of thermal images captured by drones. Using AI, temperature data and metadata (e.g., GPS position, flight altitude, camera orientation) are evaluated to identify signs of fire. The data is transferred to a cloud API and visualized in a web frontend, which also includes a configurable alarm system.
Our tasks
- Plan and test optimized flight routes
- Ensure image quality through optimal overlap
- Data transfer & real-time processing
- Develop an interface for displaying images and results