Opportunistic Memory

In this paper, we present a model of oppor-tunistic planning that uses planning-time rear soning about the opportunities that might arise during plan execution. The model is composed of three parts: a planning-time mechanism that places blocked goals into memory for later activation, an understanding system that activates suspended goals as a by-product of parsing the world, and an execution-time process for evaluating opportunities and merging newly activated goals into the plan-ning/execution agenda. We discuss this model in terms of examples from TRUCKER, a route-scheduling system, and RUNNER, an errand planner. Current research in planning has taken a rather dramatic change in course in the past few years. The notion that a planner can exhaustively preplan for a set of goals prior to execution has been largely abandoned. In part, this change is the result of demonstrations that planning for conjunctive goals is undecidable [Chapman, 1987]. A more important factor has been the realization that the dual assumptions of traditional planning (a closed world and complete knowledge) are untenable in any but the simplest domains. The planning theories that have grown out of this shift in paradigm differ from earlier theories in that they attempt to integrate planning and execution into a single process. The most extreme example of this has been the idea of situated activity [Agre & Chapman, 1987], which argues that plan-like behavior rises out of reflexive responses to external cues rather than from the guidance of a declarative plan. Less extreme theories center around the notion that plans have to be refined and repaired at execution time. These theories include: • Alterman's Adaptive Planning [1985], in which execution-time problems are handled by moving between "semantically" similar plan steps. • Firby's RAPs [1987], which allow a hierarchical planner to select between alternative plans on the basis of bottom-up information obtained at execution time. • Georgeff and Lansky's Procedural Reasoning System [1987], in which changing goals and beliefs about the state of the world determine which plans are chosen to be put on the execution stack. • Hammond's Case-Based Planning [1989] and Sim-mon's and Davis' Generate, Test, and Debug [1987], both of which use causal explanations of execution-time failures to choose between a variety of repairs. Moderate or extreme, each of these theories argues that the world changes, and that a planner must respond to those changes. They define a class of planners which, for lack of a …