This research introduces an intelligent, distributed approach for predictive scheduling and reactive scheduling. Product descriptions and their manufacturing requirements are represented using a feature-based modeling scheme. Manufacturing resources, including facilities and persons, are modeled as agents. In predictive scheduling, the optimal production schedule is identified through heuristic search and agent-based negotiation. In reactive scheduling, the original schedule is modified for responding to the changes of orders and resources through a match-up approach and agent-based negotiation, when the original schedule cannot be accomplished due to these changes.