Despite recent progress in AI planning, many benchmarks remain challenging for current planners. In many domains, the performance of a planner can greatly be improved by discovering and exploiting information about the domain structure that is not explicitly encoded in the initial PDDL formulation. In this paper we present and compare two automated methods that learn relevant information from previous experience in a domain and use it to solve new problem instances. Our methods share a common four-step strategy. First, a domain is analyzed and structural information is extracted, then macro-operators are generated based on the previously discovered structure. A filtering and ranking procedure selects the most useful macro-operators. Finally, the selected macros are used to speed up future searches. We have successfully used such an approach in the fourth international planning competition IPC-4. Our system, Macro-FF, extends Hoffmann’s state-of-the-art planner FF 2.3 with support for two kinds of macro-operators, and with engineering enhancements. We demonstrate the effectiveness of our ideas on benchmarks from international planning competitions. Our results indicate a large reduction in search effort in those complex domains where structural information can be inferred.

LAMA is a classical planning system based on heuristic forward search. Its core feature is the use of a pseudo-heuristic derived from landmarks, propositional formulas that must be true in every solution of a planning task. LAMA builds on the Fast Downward planning system, using finite-domain rather than binary state variables and multi-heuristic search. The latter is employed to combine the landmark heuristic with a variant of the well-known FF heuristic. Both heuristics are cost-sensitive, focusing on high-quality solutions in the case where actions have non-uniform cost. A weighted A ∗ search is used with iteratively decreasing weights, so that the planner continues to search for plans of better quality until the search is terminated. LAMA showed best performance among all planners in the sequential satisficing track of the International Planning Competition 2008. In this paper we present the system in detail and investigate which features of LAMA are crucial for its performance. We present individual results for some of the domains used at the competition, demonstrating good and bad cases for the techniques implemented in LAMA. Overall, we find that using landmarks improves performance, whereas the incorporation of action costs into the heuristic estimators proves not to be beneficial. We show that in some domains a search that ignores cost solves far more problems, raising the question of how to deal with action costs more effectively in the future. The iterated weighted A ∗ search greatly improves results, and shows synergy effects with the use of landmarks. 1.

Despite recent progress in AI planning, many problems remain challenging for current planners. In many domains, the performance of a planner can greatly be improved by discovering and exploiting information about the domain structure that is not explicitly encoded in the initial PDDL formulation. In this paper we present an automated method that learns relevant information from previous experience in a domain and uses it to solve new problem instances. Our approach produces a small set of useful macro-operators as a result of a training process. For each training problem, a structure called a solution graph is built based on the problem solution. Macro-operators with partial ordering of moves are extracted from the solution graph. A filtering and ranking procedure selects the most useful macro-operators, which will be used in future searches. We introduce a heuristic technique that uses only the most promising instantiations of a selected macro for node expansion. Our results indicate an impressive reduction of the search effort in complex domains where structure information can be inferred.

We present a novel approach to learning heuristic functions for AI planning domains. Given a state, we view a relaxed plan (RP) found from that state as a relational database, which includes the current state and goal facts, the actions in the RP, and the actions ’ add and delete lists. We represent heuristic functions as linear combinations of generic features of the database, selecting features and weights using training data from solved problems in the target planning domain. Many recent competitive planners use RP-based heuristics, but focus exclusively on the length of the RP, ignoring other RP features. Since RP construction ignores delete lists, for many domains, RP length dramatically under-estimates the distance to a goal, providing poor guidance. By using features that depend on deleted facts and other RP properties, our learned heuristics can potentially capture patterns that describe where such under-estimation occurs. Experiments in the STRIPS domains of IPC 3 and 4 show that best-first search using the learned heuristic can outperform FF (Hoffmann & Nebel 2001), which provided our training data, and frequently outperforms the top performances in IPC 4.

The availability of informed (but inadmissible) planning heuristics has enabled the development of highly scalable planning systems. Due to this success, a body of work has grown around modifying these heuristics to handle extensions to classical planning. Most recently, there has been an interest in addressing partial satisfaction planning problems, but existing heuristics fail to address the complex interactions that occur in these problems between action and goal selection. In this paper we provide a unique heuristic based on linear programming that we use to solve a relaxed version of the partial satisfaction planning problem. We incorporate this heuristic in conjunction with a lookahead strategy in a branch and bound algorithm to solve a class of over-subscribed planning problems.

The treatment of exogenous events in planning is practically important in many realworld domains where the preconditions of certain plan actions are affected by such events. In this paper we focus on planning in temporal domains with exogenous events that happen at known times, imposing the constraint that certain actions in the plan must be executed during some predefined time windows. When actions have durations, handling such temporal constraints adds an extra difficulty to planning. We propose an approach to planning in these domains which integrates constraint-based temporal reasoning into a graph-based planning framework using local search. Our techniques are implemented in a planner that took part in the 4th International Planning Competition (IPC-4). A statistical analysis of the results of IPC-4 demonstrates the effectiveness of our approach in terms of both CPU-time and plan quality. Additional experiments show the good performance of the temporal reasoning techniques integrated into our planner. 1.

Abstract. Current efficient planners employ an informed search guided by a heuristic function that is quite expensive to compute. Thus, ordering nodes in the search tree becomes a key issue, in order to select efficiently nodes to evaluate from the successors of the current search node. In a previous work, we successfully applied a CBR approach to order nodes for evaluation, thus reducing the number of calls to the heuristic function. However, once cases were learned, they were not modified according to their utility on solving planning problems. We present in this work a scheme for learning case quality based on its utility during a validation phase. The qualities obtained determine the way in which these cases are preferred in the retrieval and replay processes. Then, the paper shows some experimental results for several benchmarks taken from the International Planning Competition (IPC). These results show the planning performance improvement when case utilities are used. 1

Abstract. One of the most successful approaches in automated planning is to use heuristic state-space search. A popular heuristic that is used by a number of state-space planners is based on relaxing the planning task by ignoring the delete effects of the actions. In several planning domains, however, this relaxation produces rather weak estimates to guide search effectively. We present a relaxation using (integer) linear programming that respects delete effects but ignores action ordering, which in a number of problems provides better distance estimates. Moreover, our approach can be used as an admissible heuristic for optimal planning.

A number of today’s state-of-the-art planners are based on forward state-space search. The impressive performance can be attributed to progress in computing domain independent heuristics that perform well across many domains. However, it is easy to find domains where such heuristics provide poor guidance, leading to planning failure. Motivated by such failures, the focus of this paper is to investigate mechanisms for learning domain-specific knowledge to better control forward search in a given domain. While there has been a large body of work on inductive learning of control knowledge for AI planning, there is a void of work aimed at forward-state-space search. One reason for this may be that it is challenging to specify a knowledge representation for compactly representing important concepts across a wide range of domains. One of the main contributions of this work is to introduce a novel feature space for representing such control knowledge. The key idea is to define features in terms of information computed via relaxed plan extraction, which has been a major source of success for non-learning planners. This gives a new way of leveraging relaxed planning techniques in the context of learning. Using this feature space, we describe three forms of control knowledge—reactive policies (decision list rules and measures of progress) and linear heuristics—and show how to learn them and incorporate them into forward state-space search. Our empirical results show that our approaches are able to surpass state-of-the-art nonlearning planners across a wide range of planning competition domains.

We present a heuristic search approach to solve partial satisfaction planning (PSP) problems. In these problems, goals are modeled as soft constraints with utility values, and actions have costs. Goal utility represents the value of each goal to the user and action cost represents the total resource cost (e.g., time, fuel cost) needed to execute each action. The objective is to find the plan that maximizes the trade-off between the total achieved utility and the total incurred cost; we call this problem PSP NET BENEFIT. Previous approaches to solving this problem heuristically convert PSP NET BENEFIT into STRIPS planning with action cost by pre-selecting a subset of goals. In contrast, we provide a novel anytime search algorithm that handles soft goals directly. Our new search algorithm has an anytime property that keeps returning better quality solutions until the termination criteria are met. We have implemented this search algorithm, along with relaxed plan heuristics adapted to PSP NET BENEFIT problems, in a forward state-space planner called Sapa PS. An adaptation of Sapa PS, called Yochan PS, received a “distinguished performance” award in the “simple preferences” track of the 5 th International Planning Competition.