The effectiveness of a control room relies heavily on the expertise of the staff working under difficult and stressful conditions. The increased workload, that arises as complex situations unfold, adversely affects the quality of decision-making leading to reduced effectiveness. Significant benefits and improvements in performance can be achieved from deployment of advanced decision support systems (DSS) to aid control staff in carrying out their duties. A DSS is designed to co-operate with the staff's own decision-making and assist them in detailed implementation of their chosen strategies. The potential of DSS will be illustrated through a prototype system designed to aid Metro line-control staff recover normal service following an incident, e.g. broken down train. The system is designed to assist line-control staff in the detailed implementation of their chosen recovery strategy and provide specific support for crew and train resource management. Further, through forward simulation, the system is able to monitor and anticipate short- and longer-term performance issues and propose refinements to the chosen strategy. The DSS employs an agent-based framework. This is particularly suited to rail lone-control operations, where continuous co-operation and negotiation between control centre, depots and crew managers is essential to achieve effective and efficient service recovery. In our implementation, the agents are structured in layers that address distinct functional areas. A user presentation layer for anticipating users interests and requirements and aid negotiation between other users; a service layer for planning, provision and prediction of rail services; and a resource layer for monitoring, negotiating and managing resources. Advanced decision support has the capability to reduce the problem of information overload and increase control room effectiveness. Increased control effectiveness will, through improved responses to events, allow the resources under control, whether that is road or rail, to be used more effectively. For the covering abstract see ITRD E128680.
Abstract