BLACKBOARD ARCHITECTURE FOR REACTIVE SCHEDULING

This paper presents a knowledge-based approach for dynamic rescheduling. Scheduling is a complex activity often influenced by diverse and conflicting constraints. The problem is complicated further by dynamic environments, where rescheduling in response to changes is critical. In this paper a generalized architecture for dynamic rescheduling is presented. The architecture combines a blackboard-based opportunistic scheduling system that is integrated with reason maintenance capabilities. The blackboard component focuses on the process of constructing a schedule that is deemed to be primarily opportunistic in nature. The reason maintenance capabilities of the system focus on rescheduling, keeping track of justifications for assignments, and ensuring that the current schedule is always valid, that is, free of inconsistencies.