The problem of domain-independent automated planning has been a topic of research in Artificial Intelligence since the very beginnings of the field. Due to the desire not to rely on vast quantities of problem specific knowledge, the most widely adopted approach to automated planning is search. The topic of this thesis is the development of methods for achieving effective search control for domain-independent optimal planning through the construction of admissible heuristics. The particular planning problem considered is the so called “classical” AI planning problem, which makes several restricting assumptions. Optimality with respect to two measures of plan cost are considered: in planning with additive cost, the cost of a plan is the sum of the costs of the actions that make up the plan, which are assumed independent, while in planning with time, the cost of a plan is the total execution time – makespan – of the plan. The makespan optimization objective can not, in general, be formulated as a sum of independent action costs and therefore necessitates a problem model slightly different from the classical one. A further small extension to the classical model is made with the introduction of two forms of capacitated resources. Heuristics are developed mainly for regression planning, but based on principles general enough that heuristics for other planning search spaces can be derived on the same basis. The thesis describes a collection of methods, including the h^m, additive h^m and improved pattern database heuristics, and the relaxed search and boosting techniques for improving heuristics through limited search, and presents two extended experimental analyses of the developed methods, one comparing heuristics for planning with additive cost and the other concerning the relaxed search technique in the context of planning with time, aimed at discovering the characteristics of problem domains that determine the relative effectiveness of the compared methods. Results indicate that some plausible such characteristics have been found, but are not entirely conclusive.

The h(m) admissible heuristics for (sequential and temporal) regression planning are defined by a parameterized relaxation of the optimal cost function in the regression search space, where the parameter m offers a trade-off between the accuracy and computational cost of the heuristic. Existing methods for computing the h(m) heuristic require time exponential in m, limiting them to small values (m <= 2). The h(m) heuristic can also be viewed as the optimal cost function in a relaxation of the search space: this paper presents relaxed search, a method for computing this function partially by searching in the relaxed space. The relaxed search method, because it compute h(m) only partially, is computationally cheaper and therefore usable for higher values of m. The (complete) h(2) heuristic is combined with partial hm heuristics , for m = 3, ... computed by relaxed search, resulting in a more accurate heuristic. This use of the relaxed search method to improve on the h(2) heuristic is evaluated by comparing two optimal temporal planners: TP4, which does not use it, and HSP*(a), which uses it but is otherwise identical to TP4. The comparison is made on the domains used in the 2004 International Planning Competition, in which both planners participated. Relaxed search is found to be cost effective in some of these domains, but not all. Analysis reveals a characterization of the domains in which relaxed search can be expected to be cost effective, in terms of two measures on the original and relaxed search spaces. In the domains where relaxed search is cost effective, expanding small states is computationally cheaper than expanding large states and small states tend to have small successor states.

Admissible heuristics are critical for effective domain-independent planning when optimal solutions must be guaranteed. Two useful heuristics are the h^m heuristics, which generalize the reachability heuristic underlying the planning graph, and pattern database heuristics. These heuristics, however, have serious limitations: reachability heuristics capture only the cost of critical paths in a relaxed problem, ignoring the cost of other relevant paths, while PDB heuristics, additive or not, cannot accommodate too many variables in patterns, and methods for automatically selecting patterns that produce good estimates are not known.

We introduce two refinements of these heuristics: First, the additive h^m heuristic which yields an admissible sum of h^m heuristics using a partitioning of the set of actions. Second, the constrained PDB heuristic which uses constraints from the original problem to strengthen the lower bounds obtained from abstractions.

The new heuristics depend on the way the actions or problem variables are partitioned. We advance methods for automatically deriving additive h^m and PDB heuristics from STRIPS encodings. Evaluation shows improvement over existing heuristics in several domains, although, not surprisingly, no heuristic dominates all the others over all domains.

We investigate two methods of using limited search to improve admissible heuristics for planning, similar to pattern databases and pattern searches. We also develop a new algorithm for searching AND/OR graphs

In this paper, I explore the idea that there are “patterns”,analogous to software design patterns, in the kind of task proceduresthat frequently form the reactive component of architectures for intelligentautonomous systems. The investigation is carried out mainlywithin the context of the WITAS UAV project.

Planning systems rely on knowledge about the problems they have to solve: The problem description and in many cases advice on how to find a solution. This paper is concerned with a third kind of knowledge which we term domain knowledge: Information about the problem that is produced by one component of the planner and used for advice by another. We first distinguish domain knowledge from the problem description and from advice, and argue for the advantages of the explict use of domain knowledge. Then we identify three classes of domain knowledge for which these advantages are most apparent and define a language, DKEL, to represent these classes. DKEL is designed as an extension to PDDL.

The WITAS project aims to develop technologies to enable an Unmanned Airial Vehicle (UAV) to operate autonomously and intelligently, in applications such as traffic surveillance and remote photogrammetry. Many of the necessary control and reasoning tasks, e.g. state estimation, reidentification, planning and diagnosis, involve prediction as an important component. Prediction relies on models, and such models can take a variety of forms. Model design involves many choices with many alternatives for each choice, and each alternative carries advantages and disadvantages that may be far from obvious. In spite of this, and of the important role of prediction in so many areas, the problem of predictive model design is rarely studied on its own.

In this thesis, we examine a range of applications involving prediction and try to extract a set of choices and alternatives for model design. As a case study, we then develop, evaluate and compare two different model designs for a specific prediction problem encountered in the WITAS UAV project. The problem is to predict the movements of a vehicle travelling in a traffic network. The main difficulty is that uncertainty in predictions is very high, du to two factors: predictions have to be made on a relatively large time scale, and we have very little information about the specific vehicle in question. To counter uncertainty, as much use as possible must be made of knowledge about traffic in general, which puts emphasis on the knowledge representation aspect of the predictive model design.

The two mode design we develop differ mainly in how they represent uncertainty: the first uses coarse, schema-based representation of likelihood, while the second, a Markov model, uses probability. Preliminary experiments indicate that the second design has better computational properties, but also some drawbacks: model construction is data intensive and the resulting models are somewhat opaque.

We present an algorithm for planning with time and resources based on heuristic search. The algorithm minimizes makespan using an admissible heuristic derived automatically from the problem instance. Estimators for resource consumption are derived in the same way. The goals are twofold: to show the flexibility of the heuristic search approach to planning and to develop a planner that combines expressivity and performance. Two main issues are the definition of regression in a temporal setting and the definition of the heuristic estimating completion time. A number of experiments are presented for assessing the performance of the resulting planner.