Stackelberg-based Coverage Approach in Nonconvex Environments

This paper introduces StaCo: Stackelberg-based Coverage approach for nonconvex environments. This approach structurally differs from existing methods to cover a nonconvex environment, as it is based on a game-theoretic concept of Stackelberg games. Our key assumption is that one robot can predict (short-term) behavior of other robots. No direct communication takes place among the robots, the approach is decentralized. However, the leading robot can direct the system into the optimal setting much more efficiently just by changing its own position. This paper extends our previous work in which we have introduced the StaCo approach for coverage of a convex environment, with a simpler type of robots. We provide theoretical foundations of the approach. We demonstrate its benefits by means of case studies (using the Sim.I.am software). We show situations in which the StaCo approach outperforms the standard approach, which is based on combination of the Lloyd algorithm and path planning.