Optimistic and pessimistic reliability performance assessment approaches for reconfigurable manufacturing systems

• Epistemic uncertainty based reliability assessment of reconfigurable manufacturing systems • Reliability performance models built by optimistic and pessimistic assessment approaches • Numerical analysis of the optimistic and pessimistic based reliability models The importance of assessing the reliability of a complex system paradigm such as a reconfigurable manufacturing system arises from its relationship with quality, efficiency, flexibility, complexity, safety and application. However, during practical reliability analysis of a reconfigurable manufacturing system, epistemic uncertainty can emerge since the degradation information of the complex system can be insufficient, making it critical to develop a response approach. Therefore, this study develops an approach which takes into account the two extreme possibilities - optimistic and pessimistic - to set the boundary for the reliability performance of the system. The approach is proposed specifically for a reconfigurable manufacturing system with variable configuration orders. The optimistic assessment assumes that the system does not have a usage history when reconfigured with a new tool. In contrast, the pessimistic assessment assumes that the usage history has full impact on the newly reconfigured system. An algorithm is developed for the optimistic and pessimistic assessment model based on Monte Carlo simulation. The simulation result demonstrates the ability of the approach to analyze different levels of optimism and pessimism, and to quantify the uncertainty of reliability performance of a reconfigurable manufacturing system.