Model based reconfiguration of flexible production systems

Abstract Shopfloor uncertainty is a key aspect that limits the flexibility potential of nowadays manufacturing systems. Modular robotic systems are considered as a main enabler for production system reconfigurability. However, their fixed control logic, based on pre-programmed operations, does not allow the effective exploitation of their capabilities. This paper proposes a scalable assembly execution control framework that facilitates the real time a) process level reconfiguration and b) resource level behavior adaptation considering the dynamic changes of the working environment. A set of cognition modules for environment and process perception are wrapped as digital services consumed by the execution orchestration mechanism. A synthesis of multiple sensors’ data is created comprising a digital twin of the workplace that is real time updated based on the sensors’ input. This sensor based reconstructed scene is used as input to the cognition modules facilitating the real time reconfiguration of the manufacturing operations execution. This set of digital services has been deployed and tested in a case study from the automotive industry employing mobile dual arm workers and human operators.