Market-based Approaches for Coordination of Multi-robot Teams at Different Granularities of Interaction

Multi-robot teams can improve safety and increase human productivity for operations in hazardous environments. To be effective, a control scheme is needed to decompose a task, assign subtasks to individual robots, and synchronize execution. We have developed a market model for this control scheme that realizes the best of both centralized and distributed approaches. In the market approach, robots coordinate opportunistically to meet team constraints and to optimize the team solution. In this paper, we illustrate how the market is used to coordinate at the task decomposition, assignment, and execution phases, depending on the requirements of the given application. We present results from simulation and from actual robots for the applications of mapping, area reconnaissance, and perimeter sweeping.