Modeling Sheet Metal Integrated Production Planning for Laser Cutting and Air Bending

Over the past few years both sheet metal process planning and production planning issues received increased attention. For process planning of the laser cutting process, nesting algorithms are developed in order to decrease the waste material. Additionally, algorithms are available for path planning, i.e. determining the best sequence for cutting the different parts. Production planning is mainly performed based on the ability to fill a sheet. For air bending, process planning focuses on bend sequencing and tool selection, while production planning optimization aims at minimizing time consuming setups between the different production layouts at the press brake. However, when integrating both processes, the benefits from individual optimization counteract one another: good nestings at the laser machine can create additional setups at the press brake, hence increasing the makespan. An integrated approach is proposed to verify whether this problem can be solved by already taking into account possible setups at the press brake in the early nesting stage. Integration of both processes aims at an optimal combination of parts on a sheet and minimization of the setups at the press brake. In this paper, an overview of a modeling effort addressing both goals is proposed. When combining parts on a sheet, preference is given to parts requiring the same production layout at the press brake. If this is impossible, production layouts with low changeover times are preferred. Industrial cases are used to verify the applicability of the proposed model. The results are compared to a reference approach where nesting is performed with dedicated software and planning for air bending is based on an operator’s experience. Compared to this reference approach, a makespan reduction and a setup time reduction can be observed. The planning is generated almost instantaneously and no additional sheets are required compared to the reference approach. Future research will focus on expanding the model and verifying its applicability on a larger data-set.