Temporal networks

Integration of planning and scheduling requires new approaches to the scheduling problem. The scheduler must be able to provide useful information for the planner in order to avoid generation of unfeasible plans. In constraint-based scheduling it is possible to de ne custom ltering rules that improve the solving procedure. If the ltering rules exploit the information shared by the planner and the scheduler (e.g. precedence or temporal constraints), the outcome of these rules can be used to provide useful hints for the planner. This work presents a ltering technique that exploits temporal relations between a set of activities allocated to one or more disjunctive resources. The work also presents a set of propagation rules for constraint-based scheduling based on various ltering techniqes