An empirical study of hybrid inspection systems and allocation of inspection functions

With the demand for zero‐defect products and the requirement of shorter lead‐time production, designs of inspection systems have changed from using sampling inspection and human inspectors to 100% inspection and automated inspection systems. Unfortunately, automated inspection systems usually cannot meet the required performance without some human intervention. The objectives of this research are to demonstrate the feasibility of hybrid inspection systems, to compare these systems with the unaided human system and unaided automated inspection systems, and to develop a framework to allocate humans and computers in an inspection system in order to achieve the best system performance. This research found that a hybrid inspection system had better performance than either human inspection or entirely automated inspection systems. A proposed function allocation procedure was found to be feasible.