A Simple Testbed for On-line Planning

In contrast to most academic work on AI planning, many potential applications have an on-line character. For example, the true objective may not be to nd the fastest plan, but to have solving plus plan execution end as soon as possible. Or additional goals may arrive while previous plans are still being executed. While specic planning systems have been built to address realworld problems in such settings, the fundamental algorithmic issues have often been obscured by the myriad complexities inherent in any deployed system. Basic questions regarding the strengths and weakness of different planning approaches remain unresolved in the on-line setting. To enable the systematic study of this important area, we introduce a general simulation testbed and demonstrate its power by highlighting the critical sensitivities of two popular planning approaches when they are adapted to an on-line setting.

[1]  Edmund H. Durfee,et al.  A Survey of Research in Distributed, Continual Planning , 1999, AI Mag..

[2]  Martha E. Pollack,et al.  Introducing the Tileworld: Experimentally Evaluating Agent Architectures , 1990, AAAI.

[3]  Russell Bent,et al.  Regrets Only! Online Stochastic Optimization under Time Constraints , 2004, AAAI.

[4]  Subbarao Kambhampati,et al.  Replanning: a New Perspective , 2005 .

[5]  Maria Fox,et al.  PDDL2.1: An Extension to PDDL for Expressing Temporal Planning Domains , 2003, J. Artif. Intell. Res..

[6]  E. Onaindía,et al.  An architecture to integrate planning and execution in dynamic environments , 2003 .

[7]  J. Christopher Beck,et al.  A General Framework for Scheduling in a Stochastic Environment , 2007, IJCAI.

[8]  Martha Pollack,et al.  Benchmarks, Testbeds, Controlled Experimentation, and the Design of Agent Architectures Experimentation, and the Design of Agent Architectures , 1993 .

[9]  Subbarao Kambhampati,et al.  Improving Temporal Flexibility of Position Constrained Metric Temporal Plans , 2003, ICAPS.

[10]  Rob Sherwood,et al.  Using Iterative Repair to Improve the Responsiveness of Planning and Scheduling , 2000, AIPS.

[11]  Badr Al-Badr.  Critiquing the Tileworld : Agent Architectures , Planning Benchmarks , and Experimental Methodology , 1991 .

[12]  Bernhard Nebel,et al.  The FF Planning System: Fast Plan Generation Through Heuristic Search , 2011, J. Artif. Intell. Res..

[13]  Wheeler Ruml,et al.  Best-First Utility-Guided Search , 2007, IJCAI.

[14]  Subbarao Kambhampati,et al.  When is Temporal Planning Really Temporal? , 2007, IJCAI.

[15]  Daniel S. Weld.  An Introduction to Least Commitment Planning , 1994, AI Mag..

[16]  Michael Pinedo,et al.  Scheduling: Theory, Algorithms, and Systems , 1994 .

[17]  Ivan Serina,et al.  Plan Stability: Replanning versus Plan Repair , 2006, ICAPS.

[18]  Malte Helmert,et al.  The Fast Downward Planning System , 2006, J. Artif. Intell. Res..

[19]  Erann Gat,et al.  Robust Periodic Planning and Execution for Autonomous Spacecraft , 1997, IJCAI.

[20]  David E. Smith,et al.  Temporal Planning with Mutual Exclusion Reasoning , 1999, IJCAI.

[21]  Yuri Gawdiak,et al.  The Challenge of Planning and Execution for Spacecraft Mobile Robots , 2002 .

[22]  Wheeler Ruml,et al.  On-line Planning and Scheduling for High-speed Manufacturing , 2005, ICAPS.