I am working on an order picking project, which requires to cube the items into several totes in a trolley, then batching the totes, followed by parallel aisle picking/routing.

The input data has some orders with items. Items have locations, 3 dimensions. Each tote has the same size. One trolley can fit 4 totes.

The cubing part could just use approximate volume.

The goal is to create batches with planned picking route.

The size of the problem: about 50-80 items per order. 5-10 orders during the picking window.

The problem is intertwined if we want to minimize the distance. Should the cubing part solved separately without considering the batching and routing? If so, is this a VRP?

I assumed the cubing is a bin backing/multiple knapsack problem, and is first fit algorithm good enough? Which heuristic performs better?

  • $\begingroup$ This does indeed seem like it has elements of VRP, but it has other complicating elements as well. What is your question exactly? $\endgroup$ – LarrySnyder610 May 24 '19 at 23:33
  • $\begingroup$ @LarrySnyder610 please see the updates in the question $\endgroup$ – janicebaratheon May 25 '19 at 3:18
  • 1
    $\begingroup$ Your question is still very broad. If you remove the non-routing parts of the problem, then yes, it becomes a routing problem. But as far as how to decompose the problem, which heuristic to use, etc. -- those questions are impossible to answer without more details about your problem, and even then, most likely you will have to just experiment with different approaches to find what works best for your specific problem. You'll have better luck if you ask a more specific question, like, "can the first-fit heuristic be modified to solve the multiple knapsack problem?" or something like that. $\endgroup$ – LarrySnyder610 May 25 '19 at 18:14

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.