Early work on the NONLIN and O-Plan projects indicated a need for a defined methodology which would guide users performing various roles in the acquisition and analysis of domain requirements for planning. This work included links to a requirement analysis methodology, CORE (COntrolled Requirements Expression) , tool support via an intelligent assistant as part of the Task Formalism (TF) Workstation and an initial collection of guidelines and checklists to aid in using the TF domain description language. This paper describes work underway to follow-on from this past research and to infuse it with knowledge gained from recent research related to planning domain development, knowledge modelling, design rationale and ontological and requirements engineering. Introduction The activities involved in discovering, engineering, documenting, and maintaining a set of domain constructs for most AI planning-based projects can be considered ad hoc and disorganised, at best. The current sources for...

Although task reduction (HTN) planning historically preceded partial order (PO) planning, and is believed to be more general than the latter, very little comparative analysis of the two planning formalisms has been done. Part of the reason for this has been the lack of systematic understanding of the functionalities provided by HTN planning over and above that of partial order planning. In this

realistic plan representations as needed for real problem solving, and can improve the analysis that is possible for production planning systems. 2 Representing Plans as a Set of Constraints A plan is represented as a set of constraints which together limit the behaviour that is desired when the plan is executed. Work on O-Plan [8],[34] and other practical planners has identified different entities in the plan which are conveniently grouped into three types of constraint. The set of constraints describe the possible plan elaborations that can be reached or generated as shown in figure 2. Implied Constraints Plan Level Constraints Detailed Constraints Plan State Plan Agenda Plan Entities Plan Constraints<F NaN> \Gamma<F NaN> \Gamma\Psi<F NaN> @<F NaN> @R Space of Legitimate Plan Elaborations Figure 2: Plan Constraints Define Plan Space The three types of constraint in a plan are: 1. Implied Constraints or "Issues" 1 -- represe

Our research concerns formal, expressive, objectcentred languages and tools for use in engineering domains for planning applications. In this paper we extend our recent work on an object-centred language for encoding precondition planning domains to a language called OCL h , designed for HTN planning. Domain encodings for HTN planners are particularly troublesome, because they tend to be used in knowledged-based applications requiring a great deal of `domain engineering', and the abstract operators central to an HTN model do not share the fairly clear declarative semantics of concrete pre- and post condition operators. Central to our approach is the development, in parallel, of the abstract operator set and the hierarchical state specification of the objects that the operators manipulate. In this paper we define and illustrate a transparency property, together with a transparency checking tool, which helps the developer to encode a clear planning model in OCL h . Our encoding of the Tr...

Condition satisfaction in planning has received a great deal of experimental and formal attention. A "Truth Criterion" lies at the heart of many planners and is critical to their capabilities and performance. However, there has been little study of ways in which the search space of a planner incorporating such a Truth Criterion can be guided. The aim of this document is to give a description of the use of condition "type" information to inform the search of an AI planner and to guide the production of answers by a planner's truth criterion algorithm. The authors aim to promote discussion on the merits or otherwise of using such domain-dependent condition type restrictions as a means to communicate valuable information from the domain writer to a general purpose domain-independent planner 1 . Introduction to Condition Typing Research in AI planning has introduced a range of progressively more powerful techniques to address increasingly more realistic applications (Allen, Hendler & Ta...

Although task reduction (HTN) planning historically preceded partial order (PO) planning, and is understood to be more general than the latter, very little analysis has been done regarding its performance. Part of the reason for this has been the lack of systematic understanding of the functionalities provided by HTN planning over and above that of partial order planning. HTN planning has been characterized as everything from a panacea for the problems of partial order planners to a mere "efficiency hack" on partial order planning. In this paper I will extend a generalized algorithm for partial order planning, that I developed recent work, to cover HTN planning. I will use this as a basis to separate the essential and inessential differences between HTN and partial order planning.

. The paper adresses proof planning as a specific AI planning. It describes some peculiarities of proof planning and discusses some possible cross-fertilization of planning and proof planning. 1 Introduction Planning is an established area of Artificial Intelligence (AI) whereas proof planning introduced by Bundy in [2] still lives in its childhood. This means that the development of proof planning needs maturing impulses and the natural questions arise What can proof planning learn from its Big Brother planning?' and What are the specific characteristics of the proof planning domain that determine the answer?'. In turn for planning, the analysis of approaches points to a need of mature techniques for practical planning. Drummond [8], e.g., analyzed approaches with the conclusion that the success of Nonlin, SIPE, and O-Plan in practical planning can be attributed to hierarchical action expansion, the explicit representation of a plan's causal structure, and a very simple form of propo...

We present an extension to the HTN planning paradigm, which provides a more knowledge centred assessment of activities and ordering constraints. Based upon a knowledge-rich model, DART-Network planning applies mixed mode reasoning to determine the need for activities and ordering constraints within a plan. We describe a simple construction problem to demonstrate the limitations of existing HTN formalisms and the advantages of our approach. 1.

Proof planning is the application of Artificial Intelligence planning techniques to prove mathematical theorems. While exploring the domain of the residue classes over the integers with the multi-strategy proof planner Multi we found a class of hard problems on which proof planning showed a remarkable high degree of variance. On problems of the same complexity we either succeeded very quickly with short proofs or the proof planning process took significantly longer and resulted in a large proof. Recent work in Artificial Intelligence points out that the unpredictability in the running time of heuristic search procedures can often be explained by the phenomenon of heavy-tailed cost distributions. Because of the non-standard nature of these heavy-tailed cost distributions the controled introduction of randomization into the search procedures and quick restarts of the randomized procedure can eliminate heavy-tailed behavior and can take advantage of short runs. In this report,...

. So far, well-founded planning concentrates ultimately on the generation of action sequences which produce desired behaviors. But, to promote a planner's success in human assistance and support, varied representational and operational flexibility is needed. The paper argues that a compositional temporal logic framework is well-suited to reach the necessary flexibility. Planning is viewed as an inference process based on various refinements and plans work as transmitters of appropriate information about this process. Among other things nonlinear planning is explained in terms of concurrent refinements. 1 Introduction Whatever kind of planner is used, the ultimate aim of well-founded classical planning is to find a ground operator sequence, which when executed in the given initial state, will produce desired behaviors. Thereby, most techniques used have concentrated on the sub-class of behavioral constraints called the goals of attainment, i.e. a single goal state has been specified. S...