Grégoire Danoy and Sune Nielsen work in a secure room underground, a somewhat paradoxical setting, given that what they are testing are drones designed to fly in swarms. But the location reflects the sensitivity of the work.
The project, based at the University of Luxembourg, is investigating how swarms of AI-controlled drones can be used to protect a defined airspace. The research dates back to 2017, well before drone disruptions to air traffic began making headlines around the world. The timing, in hindsight, was prescient.
The team currently works with nano-drones weighing just 27 grammes, though the same principles could in theory be applied outdoors using drones of up to four kilogrammes. The catch is that the European legislative framework for deploying AI-controlled drones in open airspace does not yet exist, leaving the research confined to controlled environments for now.
The French-Danish research duo focuses specifically on the defensive application of drone swarms. Other researchers are already exploring other scenarios, including the use of kamikaze drones. Danoy and Nielsen are clear, however, that artificial intelligence may allow systems to surpass human capabilities in certain respects, but the researchers are emphatic that it is humans who must set the boundaries within which those systems operate.
