Turning theory into practice
For the four LiU students Andres Mendoza, Matheus Gr枚nstedt, Tobias Tiger, and Philip T枚rner, the competition provided an excellent opportunity to put theoretical knowledge into practice. The team members are in their second year of the Master of Science in Engineering programme in Software Engineering.
Their task during the competition was to develop an autonomous system in which multiple drones cooperate as a swarm to search a designated area. The objective was to locate a missing person in the terrain and then alert the base.
To make the drone swarm easy to operate, the students integrated an AI-based language model. This allows an operator to issue instructions in plain language, for example: 鈥淪earch for a person wearing a blue jacket in the forest north of our base.鈥� By combining conventional cameras, thermal cameras, depth cameras, and the laser-based sensing technology LiDAR, they achieved a positioning accuracy of approximately ten centimetres. The entire solution was built and tested in a simulation environment using the robotics framework ROS (Robot Operating System).
鈥淲e did not win the competition, but we are pleased with what we achieved. Everyone found it both enjoyable and highly educational to participate, and we have already signed up for future hackathons,鈥� the student team says.
Broad expertise and interdisciplinarity in the PhD student team
At the same time, an interdisciplinary team of six PhD students from Center for Heterogeneous Adaptive Swarm Systems 鈥� CHASS,聽 stepped straight into the heat of the competition. Four of the six team members had been PhD students for only two months when the competition took place, while the others had almost a year of doctoral studies behind them. The team consists of Dora Antonia Mandl, Daniel Harrysson, Sebastian Olsson, Emil Wallbom, Benjamin Sundvall, and Nikolaos Fotis.
The PhD students work across disciplines, combining expertise from several research fields, ranging from human鈥揷omputer interaction and computer science to control engineering and 3D printing of composite materials. Their challenge focused on swarm coordination 鈥� how intelligent drones can best cooperate and distribute tasks without colliding or losing track of their objective.
Unlike many other teams, the PhD students built their entire concept from scratch during the competition. Throughout the event, they received valuable feedback from mentors and experts from organisations including Saab, the European Space Agency (ESA), LEAD Business Incubator, and the Swedish Air Force Helicopter Wing.
Technology for both defence and crisis management (dual use)
The PhD students鈥� research forms part of a six-year research initiative CHASS, involving around 14 newly recruited PhD students on Link枚ping University, Lund University and 脰rebro University. The initiative is funded in part by the Swedish Foundation for Strategic Research (SSF) and responds to a changing security environment in which autonomous systems are playing an increasingly important role.
At the same time, both the students and the researchers emphasise the strong dual-use nature of the technology, meaning that it can be applied in both civilian and military contexts.
The same AI and drone systems that allocate search areas in a competition simulation could, in the future, help save lives during forest fires, searches for missing persons, floods, and other major societal crises.
