Simulation-based two stage sequencing of robotic assembly operations with deformable objects

Automatic generation of assembly sequences for products containing deformable objects, is an important step required for fully automated production. Robotic manipulation with deformable objects involves handling uncertainty related to object deformation which should be accounted for during assembly sequence planning (ASP). This work describes a two-stage sequencing planning of robotic assembly operations with deformable objects. ASP is an NP-hard problem and many heuristics have been developed for solving it. We suggest a method that improves a purely heuristic ASP by using a heuristic stage based on genetic algorithm (GA) and an additional robotic simulation stage. The method was demonstrated with products with deformable objects and promising results were achieved.