question

Jorge Manuel Garcia avatar image
0 Likes"
Jorge Manuel Garcia asked Ben Wilson commented

Picking order optimization

Hi everyone,

I was wondering if there is a way to optimize a picking route following a network navigator. I've read in the documentation that Dijkstra's algorithm is used to find the shortest path between two locations, but I'm not sure if such an algorithm is included to sort the picking list to minimize the total length travelled by the transporter. I guess that the real-world WMSs (Warehouse Management System) use specific tools such as Graph theory. Can anyone confirm me if there's a way to do it with Flexsim?

I've already read an answer in another question here at this place, which suggests the use of the distancetotravel() command to evaluate at each location the nearest location to go to. The problem is that this approach doesn't guarantee the total length to be the minimal one (I just wanted to find the minimal one, since I guess that the WMS uses this and thus the most precise simulation should consider finding this minimal route).

Thanks to all

FlexSim 16.1.2
transporterorder pickingoptimizeroute
5 |100000

Up to 12 attachments (including images) can be used with a maximum of 23.8 MiB each and 47.7 MiB total.

1 Answer

Dj Moens avatar image
3 Likes"
Dj Moens answered Ben Wilson commented

Jorge, I very much doubt that 'real-world' WMS software calculate the mathematical optimal route in all instances. In principle the problem you describe is a traveling salesman problem which is (computationally) hard to solve when a lot of positions are involved. Because of the aisle structure that exists inside a warehouse there are 'standard' heuristics in a WMS that can perform already very good (S pick, U pick and variants on that), For instance the simple "go to next nearest location" might not be mathematical optimal, but would perform very good indeed (might even perform better than the rules the WMS supports in cases).

· 1
5 |100000

Up to 12 attachments (including images) can be used with a maximum of 23.8 MiB each and 47.7 MiB total.

Jorge Manuel Garcia avatar image Jorge Manuel Garcia commented ·

Thank you @dj.moens for your reply!

Sincerely, I don't really know how algorithms are implemented in WMS products, it was just an assumption. For each order this problem is simpler than the Travelling Salesman Problem, since only certain nodes are visited each time (we can use a subpraph instead of all the nodes), but, as you said, the simple "go to next nearest location" should be an accurate enough approach regarding the simulation.

Thanks for your contribution :)

0 Likes 0 ·