Determinants of an Appropriate Degree of Autonomy in a Cyber-physical Production System

Classical productions systems are migrating step-by-step into cyber-physical production systems. The addition of much more computing power and object-bound data storage will lead to new possibilities for the advancement of autonomy in production systems. Autonomous message exchange and coordination can help to prevent quality problems (for instance wrong pairing of tool and work piece) and improve the disturbance management (for instance by faster information about current and probable disturbances). Due to the fact that nearly all improvements of existing production systems with cyber-physical systems take place in real and active manufacturing sites, on-site experiments for determining an appropriate degree of autonomy for production objects are not feasible. Therefore, a lab approach is necessary. In this contribution a hybrid lab approach to simulate various degrees of autonomy is presented [1]. The paper starts with a definition of autonomy and suggests diverse measurement methods [2]. After a short introduction into the lab concept, the results of some test runs are presented where autonomous objects perform the same production program as “dumb” production objects. Finally, an outlook for further research is given.