Vehicle Routing

Philip Kilby, Paul Shaw

Abstract: Publisher SummaryThis chapter discusses the use of constraint programming (CP) on an important industrial problem: that of constructing routes for vehicles to visit a set of customers at minimum cost. The methods aim at the movement of people and goods by road. The chapter presents several surveys indicationg a variety of problems and solution methods. An important factor in these problems is that no two companies will operate exactly alike—each has its own operational methods, and each brings their own constraints—like defined areas for drivers, vehicle/customer incompatibilities, and delivery time preferences. While traditional approaches capture the essence of the problem, a software system that is to deliver finished routes must be able to incorporate these myriad individual constraints. CP offers a valuable tool for specifying and solving real-world problems. The ability to specify and add new constraints easily is important. The advantage CP brings is a much more general method of handling the core and side constraints of routing problems. The chapter also discusses wide variety of problems classes that are seen in the literature and in the real world with the same set of variables and core constraints. This can handle capacity, time and incompatibility constraints, and various types of pickup and delivery problem. CP allows the methods developed to solve the routing aspects to be leveraged to solve the related subproblem. The chapter reviews hybrid operations research (OR) and CP methods.