Abstract This paper presents an intelligent maintenance scheduling system, designed to solve complex scheduling problems, with the flexibility and robustness necessary for use in an actual operating environment on a power distribution system. The nature and complexity of various operating constraints make direct algorithmic approaches cumbersome to use with respect to various optimization criteria. The knowledge based system introduced in this paper overcomes this problem and takes the advantage of the flexibility offered by search based techniques to produce efficient and safe schedules. This intelligent maintenance scheduling system has been developed for use on a medium sized distribution network connected to a SCADA system. It takes into consideration various operational constraints due to loading, sequencing, priority, capacity, and availability of components. Once the schedule for maintenance of components has been prepared, the expert system generates an optimum switching sequence for their isolation and for transferring to other sources all the loads which would be adversely affected due to this switching. With this knowledge based approach, the optimum schedule and switching sequence can be generated in a very short time, thus making this approach effective for on-line applications.