ED: How did the idea for Falcon Vision come about?
Anas Izaaryene: Our team, Falcon Vision, is a dedicated group of TUM students developing an autonomous unmanned aerial vehicle (UAV) to detect avalanche victims and save lives. Our drones are weather-resistant and equipped with sensors for computer vision and deep learning, among other things. In the future, they will fly in hard-to-reach mountainous areas, even in strong winds or poor visibility, to support the search and rescue of avalanche victims.
Who is the team behind it?
Our team comprises 15 students from various fields, including aviation, mechanical engineering, robotics, and engineering science. Our team members range from Bachelor's and Master's students to doctoral candidates. We bring together our diverse experience and knowledge to work on innovative solutions while continuously learning. We also have openings for marketing team members and welcome dedicated problem solvers who want to contribute to our mission.
What particular challenges are you facing in developing the drone?
At the start of the project in late 2022, we worked intensively on advanced computer vision algorithms that will enable the autonomous drone to recognize various targets in the mountains in real-time and identify people in emergency situations. The drone must withstand extreme weather conditions, which puts enormous strain on every component. Therefore, we pay close attention to all areas of production.
What is the main objective of your mission?
Recognize rescue victims, perform rescue operations quickly, and save lives: Our drone is intended to contribute to the safety and effectiveness of mountain rescue while also considering the ecological impact.
By using drones for rescue missions, costs and emissions can be reduced compared to traditional helicopter missions.
What are your next steps?
Over the next few months, we will test various components and combinations of electronics and hardware in the assembly. The focus will be on the precise localization of avalanche transceivers. The drone must adjust its flight path based on the signal strength it receives. What is currently done manually by local rescuers to locate and rescue avalanche victims will be done automatically and more efficiently by our drone. By working in conjunction with the mountain rescue service, we can locate victims more quickly. In an emergency, every minute counts towards life or death.
A testing drone will be launched this winter semester to conduct flight tests and research on autopilot and flight behavior.
We are supported within the TUM cosmos by the Professorship Big Geospatial Data Management of Prof. Martin Werner, and since November 2023, we can use the offers and infrastructure of the TUM Incubator.
Contact:
https://falconvision.org