As part of Archipelago 2025, teams competed in the “Monitoring and Ensuring Fire Safety in Forest Areas” category, co-organized by Brain Development LLC. Eight teams took part, from the Republic of Buryatia, Vladimir Oblast, Moscow, Ryazan Oblast, Saint Petersburg, Sverdlovsk Oblast, Tomsk Oblast, and Voronezh Oblast.
Over three days, participants attended briefings and safety instruction, created their own datasets, and trained neural networks in NNTrack software to detect target objects—campfires in forest areas. The teams photographed the flight zones, configured autonomous drone flights, tested different neural network architectures and integration approaches, selected the most effective solutions, and presented them in scored flights.
One team completed the challenge in full, while the others completed it in part. Every team trained neural networks and demonstrated a high level of target detection. Most teams also successfully programmed their drones to fly autonomously in the designated zone, demonstrating that their solutions were ready for practical use.
Competition winners:
- Buran (Saint Petersburg)
- Flynetics (Voronezh)
- Fixiki (Moscow)
The organizers praised the teams’ preparation and successful work. They highlighted the participants’ contribution to developing drone technologies for monitoring and fire safety in forests.


