Investigation into inspection system utilisation for advanced manufacturing systems.

Abstract

Varied inspection is an aperiodic inspection utilisation methodology that was developed for advanced manufacturing systems. The inspection scheme was created as a solution to improve manufacturing performance where inspection hinders production, such as cases where inspection time is significantly larger than machining time. Frequent inspection impedes production cycles which result in undesirable blocking, starving, low machine utilisation, increased lead time and work-in-process. The aim of the inspection strategy was to aid manufacturing metrics by adjusting inspection utilisation through multiple control methods. The novelty of the research lies in using an inspection strategy for improved manufacturing performance. Quality control was traditionally viewed as an unintegrated aspect of production. As such, quality control was only used as a tool for ensuring certain standards of products, rather than being used as a tool to aid production. The problem was solved by using the amount of inspection performed as a variable, and changing that variable based on the needs of the manufacturing process. “Inspection intensity” was defined as the amount of inspection performed on a part stream and was based on inputs such as part quality, required production rates, work-in-process requirements among other factors. Varied inspection was executed using a two-level control architecture of fuzzy controllers. Lower level controllers performed varied inspection while an upper level supervisory controller measured overall system performance and made adjustments to lower level controllers to meet system requirements. The research was constrained to simulation results to test the effects of varied inspection on different manufacturing models. Simulation software was used to model advanced manufacturing systems to test the effects of varied inspection against traditional quality control schemes. Matlab’s SimEvents® was used for discrete-event simulation and Fuzzy Logic Toolbox® was used for the controller design. Through simulation, varied inspection was used to meet production needs such as reduced manufacturing lead time, reduced work-in-process, reduced starvation and blockage, and reduced appraisal costs. Machine utilisation was increased. The contribution of the research was that quality control could be used to aid manufacturing systems instead of slowing it down. Varied inspection can be used as a flexible form of inspection. The research can be used as a control methodology to improve the usage of inspection systems to enhance manufacturing performance.