Human error prediction in man-machine system using classification scheme of human erroneous actions

Abstract

While the advanced technology improves the reliability of technological systems, the reliability of human operators still remains limited. From the viewpoint of systems safety and reliability, the possibility of human error occurrence must be reduced. At the design stage of a man-machine system, it is important to predict what kind of interactions will lead to a serious accident. This paper presents a prediction method using the classification scheme of human erroneous actions, which shows the general result obtained in such fields as psychology and cognitive engineering. The human behavior cannot be described in the same way as the technological system behavior. The context dependency is one of the characteristics of human behaviors. The context determines the most probable human, erroneous action and specifies man-machine interactions in the system. A simple illustrative example shows the details and merits of the proposed method.