Welcome to Swarm-Rescue

Swarm-Rescue is a competition organized by the Interdisciplinary Center for Defense and Security (CIEDS) at IP Paris and by the Defense Innovation Agency (AID). It is open to students from IP Paris (Ecole Polytechnique, Telecom Paris, Telecom Sud Paris, ENSTA Paris, ENSAE) and the schools of the French Ministry of Defense (ISAE Supaéro, ENSTA Bretagne).

Your job will be to propose your own version of the controller of the drones. In a competition, each participating team will perform on a new unknown map, and the winner will be the one who gets the most points based on several criteria: speed, quality of exploration, number of injured people saved, number of drones returned, etc.

This challenge aims to promote among students the topics of artificial intelligence for multi-robot or multi-drone systems, an essential theme in defense and civil security. For the organizers, this challenge is also an opportunity to evaluate the long-term potential of the various technical solutions proposed by the students. Still, the organizers claim no right to the developed solutions for short-term use. All the work done remains the property of the students.

The challenge does not require any particular technical skills (beyond basic knowledge of the Python language) and will mainly mobilize creativity and scientific curiosity from the participants.

The Competition

The work on the challenge will be done exclusively in this simulation environment, with maps of increasing complexity. The drones will have to manage the limited range of communication, collaborate between them to acquire information and transmit it to an operator outside the zone, be able to manage sensor and communication failures and unforeseen events such as the loss of drones in order to conduct this mission autonomously. The final evaluation will be done on an unknown map made available to each team at the time of the demonstration.

Illustration of the swarm-rescue simulator with one drone in an empty environment.

Acknowledgments

We would like to thank the team behind Simple-Playgrounds (SPG), the simulator used to implement the competition.