Situated and Event Discrete Decision Making Support System Applied to Remotely Operated Vehicles

The monitoring of the human-machine interaction, the planning and prediction of possible behaviors, and the detection of missing actions or other errors in advance contributes to the safety of human-machine systems. The development of supervision support algorithms may focus to decision making and also to the related detection of critical decision making situations.With the knowledge of consequences the effects of possible human errors can be evaluated in advance. Based on this online generated knowledge about possibly upcoming consequences the humanmachine interaction can be effected for example by additional warnings. This is of special relevance for vehicles which are remotely operated. In the public-supported project FernBin beside others the supervision of the captain's actions in inland shipping is addressed. The focus is to model possible human-guided driving maneuvers several action steps ahead, the detection of human errors, and also of non- optimal behaviors by defining optimal ones. The final system supervises the captain's behavior supports her or him in critical situations based on the automated decision making support system.

Technical core of this contribution is a Situation-Operator-Modeling (SOM) as event-discrete approach used to model the captain-vessel-interaction of a remotely guided vessel as a graph-based-model. Using this approach sequences of possibly connected actions can be generated describing the human interaction options and therefore possibly upcoming future behaviors which allows beside the detection of not allowed actions, omitted but required actions, the detection of intended as well as unintended upcoming future situations. The approach is applied to experimentally-generated real situations within the context of the FernBin project in combination with the research vessel `Ernst Kramer'.

Keywords: Situated monitoring, Decision making support system, Human-Machine-interaction, Situation-Operator-Modeling, Cognitive technical systems.

Download PDF