• Costs. The state models discussed above can be extended to take into account action costs. Traditionally, the focus in planning has been on uniform action costs, however, costs may depend on both actions and states. For example, a fluent t standing for time can be included so that actions increase t by their time duration. Then, defining c(a , s) as the duration of a ins, i.e., c(a,s) = (tSa"-tS ) where sa = next( a, s), a formulation would be obtained in which optimal plans stand for plans with minimum completion times. In a different way, optimal parallel plans, such as those computed by Graphplan and SAT planners, can be defined as well. The issue is how to plan effectively in such flexible cost structures.