Minimization of knives setup times in the context of cutting problems where input rolls have defective areas – OMP
|Company Name / Department
Optional remote work
|Travel expenses (own account or reimbursed by the company)
|Reimbursed by the company
|Housing arranged by company
|Housing expenses (how much per month, own account or subsidized by the company)
|€450,- per month
Solving cutting and trimming plans is an essential part of OMP’s offering for the paper and plastic film industries. On some machines the time to setup individual slitting knives is time consuming and should be minimized. We already have an algorithm to do this integrated to our offering however it could probably be improved when the material being cut has defects which influences the possible positions of the products being cut. The focus of this thesis would be on defining, implementing, and benchmarking optimization methods to solve the knives setup time minimization problem in the context of cutting problems where input rolls have defective areas.
Using a top view, the cutting pattern being executed on the previous example could be represented this way (where each rectangle represents one of the three rolls being simultaneously slitted):
The graphic representation below shows three cutting patterns producing some products (represented by the blue rectangles) being cut from input material having some defects (represented by the red rectangles). The yellow dots represent the positions where a slitting knife needs to be setup. The graphic representation must be read from top to bottom:
- First, we cut two products on an input roll having two defects on the sides.
- Then we cut two products on an input roll having one narrow defect on the left side and another in the middle
- Finally, we cut one product on an input roll having one large defect on the left side.
The knife minimization problem definition is then the following:
Given some patterns each indicating how many products with some given dimensions should be produced and on which section of input roll they should be produced, decide in which order the patterns should be executed and what should be the actual position of each product within its pattern (so that it does not intersect with an incompatible defect), in order to minimize the total number of knives that will need to be setup.
Goals of the project:
The objectives of the project will be to propose an optimization method capable of producing in short times solutions of good quality, to implement it and to benchmark it on some test problems.
More information: email@example.com