Virtual Commissioning and Symbolic Planning of Micro-Optical Assembly Processes

Automation in micro-optical assembly is an emergent, but steadily growing field of research. To advance the optics industry's efforts in this regard, several methods that aid optics engineers with development and deployment of assembly processes have been proposed. However, it is still a non-trivial and time-consuming task to design efficient work orders that consider movement characteristics of integrated robotic assembly systems. In this contribution, we present an approach to combine well-investigated symbolic planning algorithms, semantic modelling techniques and 3D simulation-based virtual commissioning to generate optimized work orders for automated micro-optical assembly and thereby help to overcome the difficulties of manual process design. Our approach is a good example for the efficiency of the eRobotics concept and demonstrates how its systematic application can resolve practical problems with existing approaches. Results indicate that our approach to automated planning is, by enabling a formal definition of quality metrics for action sequences, not only faster but also usually superior to manual generation of work orders.