Task Allocation Methods for Multi-Robot Systems

Task Allocation Methods for Multi-Robot Systems

Project type: PhD project

Period: -0001 - -0001

Design of new methods of task allocation for multi-robot systems.

It is well known that multi-robot systems can provide some advantages over single robots: robustness, flexibility and efficiency among others. To benefit from these potential aspects the robots must cooperate to carry out a common mission. Among typical problems for multi-robot systems, we centre our work on task allocation aspects, that is, selecting the best robots to carry out different tasks. When forming task-groups the number of robots has an important impact on the system performance, due, for example, to the interference effect, which is the result of competition for the shared space. Therefore, a good task allocation mechanism must decide on the best number of robots needed to carry out each task, provided priorities and deadlines will be taken into account. The methods we are working on are mainly inspired on auction and threshold mechanisms.Some of the topics we work on are: teams formation, leader to leader negotiation, heterogeneous teams formation, and deadlines influence, physical interference impact modelling.To test the developed methods we have some facilities in our lab, which include, a special purpose simulator (RoboCoT) and a five Pioneer 3DX team.

Project Leader

Project Collaborators

Related Publications