An approximate dynamic programming approach to the Dynamic Traveling Repairperson Problem

This paper presents a novel suboptimal policy for the Dynamic Traveling Repairperson Problem (m-DTRP), a problem requiring dynamic planning for a team of vehicles. The suggested policy is adaptive, locally distributed, computationally efficient, and independent of traffic load intensities. It is shown that the policy is asymptotically optimal in the light traffic load case. Experimental results are provided to show that the performance in the moderate and heavy load cases is comparable to the best known policies.