Solving VRPTW with delay route to satisfy clustering constraint

This paper proposes a new area of research for the widely researched Vehicle Routing Problem with Time Window constraints, VRPTW. The problem was redefined with an additional constraint which is to be called the clustering constraints. The reason behind adding this constraint was to realistically save the business goal, which is the real-world operating cost of a logistics company. A proposed solution that can cater for this constraint in solving VRPTW problems was introduced. The solution was to add additional “delay routes” in order to plan those orders that cannot be fitted in today's route simulating the predicted tomorrow's route based on the unplanned orders for today. In our experiments the benchmark from [2] Solomon Marius M's VRPTW Benchmark Problems as well as [3] Gehring and Homberger's extended VRPTW instances of the original Solomon benchmark datasets was utilized. We did some modification to the datasets to be able to run our experiments. We collected different results of the solution with different tuning on the prioritization of the algorithm on the new constraint.