A shared workspace to support man-machine reasoning: application to cooperative distant diagnosis

摘要

This paper focuses on man-machine cooperation problems. In particular, it deals with those problems that occur when both human and machine have to achieve a shared reasoning activity. It puts forward a man-machine approach that is dedicated to technical diagnosis problem solving. Coordination of human and automated reasoning is key to solving this problem, since efficiency depends on both sharing and interpreting exchanged data. A shared workspace is proposed to support both machines and their human operators. This workspace is kept as close as possible to human representations in order to reduce cooperation costs. The paper describes those coordination mechanisms that are able to support such a cooperative activity using a shared workspace. In order to assess the costs and benefits of such cooperation, these mechanisms are applied to a complex industrial problem: diagnosis and troubleshooting in a phone network. The results show the full impact of cooperation on human-machine reasoning.