About future improvements

[rlacjfjin]

Hi,

first of all, thanks a lot for sharing.
Regarding the improved part you mentioned in your report, is it in the work plan in the near future?
I would like to know how to make the project support 3D rotation, can you give me some advices?

Thanks,
Zhe.

[Wadaboa]

Hi,

Thanks for opening the issue. Unfortunately, we currently do not have any plans to work on "future improvements" mentioned in the slides. Will keep you posted in this issue in case we decide to extend the current code-base.

As for supporting rotation, you can easily extend the current approach to support rotation on the z axis, by introducing boolean variables in the subproblem (0 meaning original position and 1 meaning 90 degree rotation). Supporting rotation on the x and y axes is more complicated, as that would imply possibly dealing with layers as well (since a rotation might imply a change in height). Anyways, supporting full rotations in a 3D bin packing problem is not straightforward, as it introduces many more decision variables and makes the problem even harder to solve; on top of that, we worked with superitems and layers, which further complicates integration with rotation. Still, if you want to give it a shot, you can start reading up this paper for example.

Let me know if you decide to go ahead with the implementation of rotation, it would be a nice addition to the repo.

Thanks,
Alessio.

[rlacjfjin]

Thanks for your reply.
I'll look at your code first, and try to do what it does if I can.

Zhe

[raucal]

I am also trying to achieve this goal. I do the maxrects with the rotation argument equal to True and I can find out if any cja can be placed on a lower stack if we rotate it. However, I cannot then exchange the width and depth values with each other since I get an error that the value of these attributes cannot be directly set.