A Mixed-Integer Programming Approach for Complete Coverage Task Allocation of Heterogeneous Robots with Startup Time Cost

This paper addresses the problem of complete coverage task allocation for a heterogeneous robot team operating in a mapped workarea like an indoor building. The robots in the team have varying execution speeds, and the startup cost of robots’ traversing from the depot to its nearest allocated regions is also considered. We present a new allocation approach utilizing Mixed-Integer Programming (MIP) to tackle the heterogeneity and startup requirements. The solution of this approach is featured threefold: balanced task allocation; connectivity of the allocated regions to facilitate a safe robot execution without interference to other robots; disjoint task allocation. We present a case study showcasing the practicability of this approach in the robotic complete coverage scenario. And we further studied the performance of our approach in mapped workareas for different numbers of heterogeneous robots and demonstrated its effectiveness. Notably, this approach can successfully optimally allocate instances with a combinatorial space of 2
¹⁰⁰
within a reasonable time.