Layout Optimization of Robotized Production Systems using a Mass-Spring-Damper Model

Abstract

The shift towards more customized products presents production planners with the challenge of frequently reconfiguring their layouts to accommodate producing/manufacturing new product variants. Due to the vast solution space and time constraints, planners often default to traditional patterns when designing new layouts rather than thoroughly exploring the optimal solutions. While meta-heuristics such as genetic algorithms (GAs) can effectively navigate solution spaces, they often produce suboptimal results and require careful parameter tuning. This paper introduces an innovative method developed at Fraunhofer IPA for optimizing layouts in robotized production systems, where robots serve as handling resources to transfer parts between various operations. This approach models the layout as a Mass-Spring-Damper system. In this model, volume bodies represent the resources to be organized within the layout, while springs and dampers signify the "closeness" relationships between these resources. Friction forces are used to depict the effort required to rearrange existing resources. The optimal layout, in terms of minimizing handling distances and area usage, is achieved when the system reaches equilibrium—where the total potential energy in all springs and dampers is minimized. The approach is validated using a small greenfield layout problem involving five resources and is compared against the best layout solution derived from a genetic algorithm.