Abstract. The similarity rules used in previous case-based planning systems are not accurate enough. Consequently, any case-based planning system rarely outperforms the new planning approach called heuristic search planner, which performance has been increasing over the last years. This paper examines the use of a new and more accurate similarity rule, called ADG, in a case-based planning system, and confronts its results with the results from the FF heuristic search planner. In addition, some experiments show that in hard situations the case-based planning with ADG similarity can be better than the FF planner.
