Most planning problems have strong structures. They can be decomposed into subdomains with causal dependencies. The idea of exploiting the domain decomposition has motivated previous work such as hierarchical planning and factored planing. However, these algorithms require extensive backtracking and lead to few efficient general-purpose planners. On the other hand, heuristic search has been a successful approach to automated planning. The domain decomposition of planning problems, unfortunately, is not directly and fully exploited by heuristic search. We propose a novel and general framework to exploit domain decomposition. Based on a structure analysis on the SAS+ planning formalism, we stratify the sub-domains of a planning problem into dependency layers. By recognizing the stratification of a planning structure, we propose a space reduction method that expands only a subset of executable actions at each state. This reduction method can be combined with state-space search, allowing us to simultaneously employ the strength of domain decomposition and high-quality heuristics. We prove that the reduction preserves completeness and optimality of search and experimentally verify its effectiveness in space reduction.

Search methods based on Monte-Carlo simulation have recently led to breakthrough performance improvements in difficult game-playing domains such as Go and General Game Playing. Monte-Carlo Random Walk (MRW) planning applies Monte-Carlo ideas to deterministic classical planning. In the forward chaining planner Arvand, Monte-Carlo random walks are used to explore the local neighborhood of a search state for action selection. In contrast to the stochastic local search approach used in the recent planner Identidem, random walks yield a larger and unbiased sample of the search neighborhood, and require state evaluations only at the endpoints of each walk. On IPC-4 competition problems, the performance of Arvand is competitive with state of the art systems.

In this work we aim to narrow the gap between plan recognition and planning by exploiting the power and generality of recent planning algorithms for recognizing the set G ∗ of goals G that explain a sequence of observations given a domain theory. After providing a crisp definition of this set, we show by means of a suitable problem transformation that a goal G belongs to G ∗ if there is an action sequence π that is an optimal plan for both the goal G and the goal G extended with extra goals representing the observations. Exploiting this result, we show how the set G ∗ can be computed exactly and approximately by minor modifications of existing optimal and suboptimal planning algorithms, and existing polynomial heuristics. Experiments over several domains show that the suboptimal planning algorithms and the polynomial heuristics provide good approximations of the optimal goal set G ∗ while scaling up as well as state-of-the-art planning algorithms and heuristics.

In many real world planning domains, some observation information is optional and useless to the execution of a plan; on the other hand, information acquisition may require some kind of cost. The problem of observation reduction for strong plans has been addressed in the literature. However, observation reduction for plans with contexts (which are more general and useful than strong plans in robotics) is still a open problem. In this paper, we present an attempt to solve the problem. Our first contribution is the definition of structured plans, which can encode sequential, conditional and iterative behaviors, and is expressive enough for dealing with incomplete observation information and internal states of the agent. A second contribution is an observation reduction algorithm for plans with contexts, which can transform a plan with contexts into a structured plan that only branches on necessary observation information.

This paper describes how to learn Hierarchical Task Networks (HTNs) in nondeterministic planning domains, where actions may have multiple possible outcomes.  We discuss several desired properties that guarantee that the resulting HTNs will correctly handle the nondeterminism in the domain.  We developed a new learning algorithm, called ND-HTN-Maker, that exploits these properties.  We implemented ND-HTN-Maker in the recently-proposed HTN-Maker system, a goal-regression based HTN learning approach.  In our theoretical study, we show that ND-HTN-Maker soundly produces HTN planning knowledge in low-order polynomial times, despite the nondeterminism.  In our experiments with two nondeterministic planning domains, ND-SHOP2, a well-known HTN planning algorithm for nondeterministic domains, significantly outperformed (in some cases, by about 3 orders of magnitude) the well-known planner MBP using the learned HTNs.

This paper presents a new algorithm for plan recognition called ELEXIR (Engine for LEXicalized Intent Recognition).  ELEXIR represents the plans to be recognized with a grammatical formalism called Combinatory Categorial Grammar(CCG).  We show that representing plans with CCGs can allow us to prevent early commitment to plan goals and thereby reduce runtime.

Recent progress on external-memory MDP solvers has enabled optimal solutions to large probabilistic planning problems. However, PEMVI requires a human to manually partition the MDP before the planning algorithm can be applied — putting an added burden on the domain designer and detracting from the vision of automated planning.  This paper presents a novel partitioning scheme, which automatically subdivides the state space into blocks that respect the memory constraints. Our algorithm first applies static domain analysis to identify candidates for partitioning, and then uses heuristic search to generate a good partition.  We evaluate the usefulness of our method in the context of PEMVI across many benchmark domains, showing that it can successfully solve extremely large problems in each domain. We also compare the performance of automatic partitioning with previously reported results using human-designed partitions. Experiments show that our algorithm generates significantly superior partitions, which speed MDP solving and also yield vast memory savings.

While much work on learning in planning focused on learning domain physics (i.e., action models), and search control knowledge, little attention has been paid towards learning user preferences on desirable plans. Hierarchical task networks (HTN) are known to provide an effective way to encode user prescriptions about what constitute good plans. However, manual construction of these methods is complex and error prone. In this paper, we propose a novel approach to learning probabilistic hierarchical task networks that capture user preferences by examining user-produced plans given no prior information about the methods (in contrast, most prior work on learning within the HTN framework focused on learning “method preconditions”—i.e., domain physics—assuming that the structure of the methods is given as input). We will show that this problem has close parallels to the problem of probabilistic grammar induction, and describe how grammar induction methods can be adapted to learn task networks. We will empirically demonstrate the effectiveness of our approach by showing that task networks we learn are able to generate plans with a distribution close to the distribution of the userpreferred plans.

We consider the problem of planning in domains with continuous linear numeric change. Such change cannot always be adequately modelled by discretisation and is a key facet of many interesting problems. We show how a forward-chaining temporal planner can be extended to reason with actions with continuous linear effects. We extend a temporal planner to handle numeric values using linear programming. We show how linear continuous change can be integrated into the same linear program and we discuss how a temporal-numeric heuristic can be used to provide the search guidance necessary to underpin continuous planning. We present results to show that the approach can effectively handle duration-dependent change and numeric variables subject to continuous linear change.

Multi-agent path planning on grid maps is a challenging problem and has numerous real-life applications. Running a centralized, systematic search such as A* is complete and cost-optimal but scales up poorly in practice, since both the search space and the branching factor grow exponentially in the number of mobile units. Decentralized approaches, which decompose a problem into several subproblems, can be faster and can work for larger problems. However, existing decentralized methods offer no guarantees with respect to completeness, running time, and solution quality. To address such limitations, we introduce MAPP, a tractable algorithm for multi-agent path planning on grid maps. We show that MAPP has low-polynomial worst-case upper bounds for the running time, the memory requirements, and the length of solutions. As it runs in low-polynomial time, MAPP is incomplete in the general case. We identify a class of problems for which our algorithm is complete. We believe that this is the first study that formalises restrictions to obtain a tractable class of multi-agent path planning problems.