Towards flexible teamwork in behavior-based robots

Research into teamwork in autonomous robots is quickly gaining significant interest, motivated by future applications of multi-robot teams. A key challenge is to automate the control of teamwork, such that the designer can focus on the taskwork to be done. Teamwork architectures have been proposed to automate the interactions between team-members, in order to facilitate robust and speedy deployment. Existing architectures provide important teamwork services to deployed teams, such as synchronized task execution [4], and task allocation [3, 1, 5]. These allow the designer to spend most efforts on building controllers for the specific tasks to be carried out, rather than the collaboration of the robots.