• DocumentCode
    1241644
  • Title

    Multistrategy adaptive path planning

  • Author

    Goel, Ashok Kumar ; Ail, K.S. ; Donnellan, Michael W. ; De Silva Garza, Andres Gomez ; Callantine, Todd J.

  • Author_Institution
    Georgia Inst. of Technol., Atlanta, GA, USA
  • Volume
    9
  • Issue
    6
  • fYear
    1994
  • Firstpage
    57
  • Lastpage
    65
  • Abstract
    In 1990, we started the Router project, a multistrategy-strategy adaptive navigation path planner. It assumes that a mission planner has generated a specific mission plan and identified specific path planning tasks. Given a specific path planning task, it uses a combination of model based and case based methods to solve it. New versions of the Router system view strategic metacontrol as a kind of design task that takes as input a specification of a problem solving task and gives as output the specification of a virtual architecture for addressing it. One version of the system operates in simulated navigation worlds and provides a simple natural language interface. Another version is embodied in Stimpy, an autonomous mobile robot. Stimpy addresses issues in spatial navigation beyond path planning, such as plan execution and monitoring. Our goal is to describe our general framework of multistrategy adaptive path planning, and the specific design of the Router system. To focus this discussion, we report on a series of experiments with Router in simulated navigation worlds.<>
  • Keywords
    adaptive systems; mobile robots; navigation; path planning; problem solving; Router project; Stimpy; case based methods; multistrategy adaptive path planning; multistrategy-strategy adaptive navigation path planner; natural language interface; path planning tasks; plan execution; simulated navigation worlds; spatial navigation; specification; strategic metacontrol; virtual architecture; Circuits; Context modeling; Feedback; Indexing; Joining processes; Navigation; Organizing; Path planning;
  • fLanguage
    English
  • Journal_Title
    IEEE Expert
  • Publisher
    ieee
  • ISSN
    0885-9000
  • Type

    jour

  • DOI
    10.1109/64.363273
  • Filename
    363273