| Abstract: | Responses to emergencies are typically based on contingency plans. However, unexpected events can occur during the operation that affect safety and/or effectiveness of the activated response plan. Latest advances in communications and information technology can collect and transfer a large amount of data to the on-scene commander in real-time. The commander can then assess the potential impact of such events and decide if and how to revise the planned course of action to maintain safety and efficiency of the operation. This paper proposes a new paradigm for real-time decision support for emergency response - operational risk management. Emergency response is modelled as a large-scale operational system, including a human-machine real-time controller. The decision model is based on a topological graph structure, where the nodes are decisions and the arcs the activities. The attributes of the activities are expressed as ordinal preference values. The optimal course of action is the sequence of activities with the highest preference for resolving the emergency situation. The implementation of the decision model into a prototype decision support system is discussed. |
