This manual describes the syntax of PDDL, the Planning Domain Definition Language, the problem-specification language for the AIPS-98 planning competition. The language has roughly the the expressiveness of Pednault's ADL [10] for propositions, and roughly the expressiveness of UMCP [6] for actions. Our hope is to encourage empirical evaluation of planner performance, and development of standard sets of problems all in comparable notations.

Classical planning is the problem of finding a sequence of actions to achieve a goal given an exact characterization of a domain. An algorithm to solve this problem is presented, which searches a space of plan prefixes, trying to extend one of them to a complete sequence of actions. It is guided by a heuristic estimator based on regression-match graphs, which attempt to characterize the entire subgoal structure of the remaining part of the problem. These graphs simplify the structure by neglecting goal interactions and by assuming that variables in goal conjunctions should be bound in such a way as to make as many conjuncts as possible true without further work. In some domains, these approximations work very well, and experiments show that many classical-planning problems can solved with very little search. 1 Definition of the Problem The classical planning problem is to generate a sequence of actions that make a given proposition true, in a domain in which there is perfect informati...

This manual describes the syntax of PDDL, the Planning Domain Definition Language, the problem-specification language for the AIPS-98 planning competition. The language has roughly the the expressiveness of Pednault’s ADL [10] for propositions, and roughly the expressiveness of UMCP [6] for actions. Our hope is to encourage empirical evaluation of planner performance, and development of standard sets of problems all in comparable notations. 1

Most work in artificial intelligence and law has concentrated on modelling the type of reasoning done by trial lawyers. In fact, most lawyers' work involves planning -- for example, wills and trusts, real estate deals, and business mergers and acquisitions. Certain planning issues, such as the use of underspecified, or "open-textured" rules, are illustrated especially clearly in this domain. In this thesis, I set forth the characteristic features of planning in law, place it in the context of past artificial intelligence work in both law and planning, and describe CHIRON, a system that I have developed implementing my theory of open-textured planning in the domain of personal income tax law.

It is difficult to build intelligent computer-aided design (ICAD) programs using available expert system shells and AI programming languages. To build ICAD programs, tools are needed that support (a) generative search of design spaces, (b) deep search of design spaces to evaluate alternative designs, (c) simultaneous exploration of alternative designs to compare designs, (d) constraint posting and propagation, (e) knowledge-based control of inference, and (f) the representation of complex mechanical and electronic devices. Existing shells and programming languages either do not support these activities or provide only ad hoc and inefficient supporting mechanisms. We have constructed a logic programming system called FORLOG (FORward-chaining LOGic Programming) that provides well-integrated support for all of these activities. This paper presents the architecture of FORLOG and provides some simple examples of how FORLOG can be applied to constructing ICAD systems. 1

. Specifying communication routines transparently and explicitly as part of robots' plans rather than hiding them in separate modules makes robots' communication behavior more effective, efficient, and robust. It enables robot control systems to generate, reason about and revise their communication behavior. The controllers can also synchronize the robots' conversations with other actions and use control structures to make the communication behavior flexible and robust. In this paper, we extend RPL, a reactive plan language, to allow for controlling conversational actions. The additional constructs constitute an interface between RPL and conversational actions that is identical to the interface between RPL and continuous control processes such as navigation. The uniformity of the two interfaces and the control structures provided by RPL enable a programmer to concisely specify a wide spectrum of communication behavior. This paper describes how these extensions are implement...

V. 0.93 2005-09-06: Various bug fixes, in both the implementation and the manual. V. 0.9 2004-11-25: First alpha release 1 Grammars and Parsetrees This manual describes a deterministic, recursive-descent, lexicon-based parser (called “Lexiparse”). 1 Lexiparse works on a grammar organized around lexemes, to which are attached syntactic and “semantic ” processing instructions. Grammars also include specifications of how to construct lexemes from streams of characters. The parser uses the lexical and syntactic specifications to turn strings into sequences of parsetrees. The “semantics ” is expressed using internalizers that then turn parse trees into whatever data structures the syntactic structures are intended to correspond to. The parser and lexer are built on an abstraction called the generator, which represents a lazy list. At the highest level the parser can be thought of as a generator of internalized objects from a stream of characters, which is decomposed thus: chars →