Skip to content

Sustainable wood processing thanks to optimization

Together with FILL Gesellschaft m.b.H., RISC Software GmbH developed an optimization solution for the production of glulam beams. The algorithm developed for this purpose calculates the best cut and the optimal arrangement of the boards in the glulam beam in order to use the wood as efficiently as possible.

Machine and plant manufacturers want to ensure for their customers that the machine or plant in operation makes optimum use of the raw material for every possible product configuration (batch size 1). Therefore, the use of optimization methods is obvious. For FILL Gesellschaft m.b.H., RISC Software GmbH designed and implemented an optimization for machine control for the production of glue laminated beams. Glue-laminated beams are glued boards laid in layers on top of each other, which are used in timber construction and interior design because of their special properties. To produce such glue-laminated beams, the raw material, the boards, are first scanned at the entrance of the machine. Usable and unusable areas, such as chipped edges or knots, are identified. The implemented optimization algorithm then calculates for these scanned boards the current state of the machine and the given product configuration, how the boards should be cut in an optimal way and used in the gluelam in order to cut as little usable wood as possible. The sequence of the boards can be changed, and sections of boards can be buffered for later use. In this way, the glued laminated beams are built up layer by layer and the individual layers are glued together. The finished glued laminated beams are then further processed in downstream production steps.

Bild 3: Mittellagenproduktion © FILL Gesellschaft m.b.H. 

Picture: Center layer production © FILL Gesellschaft m.b.H. 

There is mathematics in it

The production plant in wood processing operates according to a specified cycle time of approx. 2 seconds. Within this cycle time, the calculation of the next processing steps and the transmission of the commands to the control must take place. However, in view of the diversity of the input material and the possible product configurations, there are countless solution variants that have to be evaluated in this short time. Thus, despite the high computing power of modern computers, it is necessary to design the optimization algorithms used specifically for the aspects of cycle time and quality in order to be able to calculate the optimum solution for the specific problem in the specified time. The optimization solution for FILL Gesellschaft m.b.H. has been successfully in operational use since the first quarter of 2019.

Bild 1: Leimbinder mit zwei fertig gestellten Lagen (0 und 1) und einer unvollständigen Lage (2). Die vertikalen, grauen Bereiche definieren Zonen, in denen keine Bretter enden dürfen. © RISC Software GmbH 

Figure: Glued laminated beams with two completed layers (0 and 1) and one incomplete layer (2). The vertical gray areas define zones where no boards may end. © RISC Software GmbH

Bild 2: Beispiel eines gescannten Bretts; die dunkel markierten Bereiche dürfen nicht verwendet werden, die hellen Bereiche können verwendet werden. © RISC Software GmbH 

Image: Example of a scanned board; the dark marked areas must not be used, the light areas can be used. © RISC Software GmbH

Project partner

Project details

  • Project short title: LeimOpt
  • Project long title: Sustainable wood processing thanks to optimization
  • Project partner:
    • FILL Gesellschaft m.b.H
  • Duration: 10/2018-05/2019


    Project management

    Dr. Michael Bögl

    Mathematical Optimization Specialist

    Read more