ABSTRACT
There are various definitions of visual inspection, but from a human perspective it can be simply described as the process of using the eye (aided or unaided) to make judgements about the condition of an item. From a scientific perspective, it is a change in the light’s properties after contact with the surface of the item. Whether this change is identified by the human eye, a vision enhancing accessory, or an automatic system does not change the key objective of visual inspection (Drury and Watson, 2002). In high value manufacturing, human visual inspection is critical, particularly for products requiring safety-critical standards compliance and quality assurance such as in the aerospace sector. For this reason organisations produce prescribed defect classifications and provide work standards and related training. However, previous field study work conducted by Cranfield University’s Industrial Psychology and Human Factors Group in aerospace manufacturing facilities has found that after training and over time operators often adapt the methods they originally learned and develop idiosyncratic techniques which, particularly in situations of ambiguity, involve a higher degree of subjective interpretation. Without a more in-depth understanding of these individual approaches to defect detection and classification, and their effectiveness, we are currently unable to advocate best practice in recommendations for training/process improvement. Moreover, when considering whether to automate visual inspection it is vital to gain a thorough understanding of the tasks currently undertaken by a human operator to ascertain feasibility of potential replacement technologies.
