Demoing a cutting stock problem, I was a bit overthinking things. If the material is cheap, the optimal cutting patterns are too complicated. Simplified cutting patterns are better even when this causes additional waste. So, I now am now using a simplistic Skyline heuristic and even then add constraints to simplify the cutting.
Now, this is really small potatoes compared to this [1]:
Now, this is really small potatoes compared to this [1]:
References
- Dominique Thiebaut, 2D-Packing Images on a Large Scale, INFOCOMP 2013 : The Third International Conference on Advanced Communications and Computation