Adaptive self-management of teams of autonomous vehicles

Abstract

Unmanned Autonomous Vehicles (UAVs) are increasingly deployed for missions that are deemed dangerous or impractical to perform by humans in many military and disaster scenarios. Collaborating UAVs in a team form a Self- Managed Cell (SMC) with at least one commander. UAVs in an SMC may need to operate independently or in sub- groups, out of contact with the commander and the rest of the team in order to perform specific tasks, but must still be able to eventually synchronise state information. The SMC must also cope with intermittent and permanent communication failures as well permanent UAV failures. This paper describes a failure management scheme that copes with both communication link and UAV failures, which may result in temporary disjoint sub-networks within the SMC. A communication management protocol is proposed to control UAVs performing disconnected individual operations, while maintaining the SMCs structure by trying to ensure that all members of the mission regardless of destination or task, can communicate by moving UAVs to act as relays or by allowing the UAVs to rendezvous at intermittent intervals. Copyright 2008 ACM.