Improving Proactive Human Behavior in Supervision of Manufacturing Systems Using Chronicles

Human-machine supervision is one of the essential fields doing fundamental research and producing applications that contribute to the dependability of complex manufacturing systems. Although many studies and supervision supports exist for the monitoring, diagnosis and decision-making/action-taking functions, prognosis has generally received less attention. Nonetheless, integrating a prognosis function seems a promising way to optimize means and performance of human-machine supervision. Clearly, the difficulties involved in identifying the functional role of prognosis in supervision and in quantifying its contribution to supervision performance could explain this lack of attention. Based on a review of the literature and an experimental approach, the research presented in this article provides a convincing foundation for stimulating interest in the integration of prognosis in the supervision process and the design of anticipative functions in the domain of human-machine supervision. The prognosis approach is to support the supervision human operator activity by an automated generation of hypotheses from a common information repository. This approach is based on the concept of “chronicle”.