A Decision Support System Prototype for Maintenance Management1

This paper presents a concept for a next generation Decision Support System (DSS) for maintenenance management. Such a system will integrate information from other systems dedicated to process control, condition monitoring, diagnosis, maintenance planning, and production planning. The main challenge is to support a maintenance planner in the composition of task schedules, where the focus is on condition-based maintenance tasks and the optimum merging of these stochastic tasks with periodic maintenance tasks and other pending tasks. A tree-based approach for decision analysis under uncertainty is presented, and a prototype implementing the methodologies is described. The prototype is developed using an agent-based approach for integrating distributed networked systems. The presented work is part of a strategic research project funded by the Norwegian Research Council, 2 oil companies, and MARINTEK.