| Inspection is an important step in ensuring product quality, and the goal of a defect-free product may mean that 100% of the units manufactured or serviced must be inspected to ensure consistent quality. This inspection has traditionally been conducted by either humans or machines working alone. However, neither purely human nor purely automated systems may fully achieve the desired performance in an inspection task. It is possible, though, that superior performance could be achieved by a system in which certain machine characteristics are depended upon to complement the strengths of human inspectors. One challenge in designing such a hybrid inspection system is determining how to best allocate the functions between humans and machines. Hence, there is a need to develop a methodology for using humans and machines in combination and to demonstrate the feasibility of such a hybrid system. In response to this need, a hybrid inspection simulator was developed to facilitate the conducting of controlled studies to address issues related to human machine system design and function allocation in visual inspection. Two experiments were then conducted involving the inspection of printed circuit boards using the simulator, the first, a best system performance assessment and the second, the measurement of trust in a hybrid inspection system. Results from the first experiment indicated that a hybrid inspection system where both humans and computers shared search and decision-making tasks yielded the best system performance. Results from the second showed that the trust components as well as overall trust were sensitive to different types of inspection systems, indicating that measuring trust in a hybrid inspection system is feasible. Most importantly, human adaptability was once again demonstrated, emphasizing that humans will continue to play a vital role in the design of future inspection systems. |
