A Hybrid Approach to Nonconformance Tracking and Recovery

Abstract

The nonconformance diagnosis problem has been a major issue facing industry and academia over the years. Research has been carried out on technologies for different aspects of nonconformance diagnosis such as nonconformance monitoring, prediction, prevention, classification, tracking, and recovery. Despite these advances, nonconformance tracking and recovery still receive many concerns due to the fact that they are knowledge intensive and experience-based tasks, which in complex manufacturing environments can sometimes be beyond the capabilities of skilled operators and engineers. In addition, the existing systems for nonconformance tracking and recovery are usually special purpose systems. They lack the capabilities to migrate to new working domains. This paper proposes a generic intelligent nonconformance tracking and recovery (GINTR) system. In conjunction with computational intelligent techniques such as Artificial Neural Networks (ANN) and Genetic Algorithm (GA), the system identifies the root causes of a nonconformance and provides timely corrective actions. The drive towards designing such a system is motivated by the need to implement a generic base of system capabilities that is reliable, economical, scalable, and provides a stable foundation for migrating the system to different domains.