Class #5: Sampling-Based Motion Planning: Probabilistic Roadmaps



The figure on the left illustrates the concept of a probabilistic roadmap. The figure in the middle shows the milestones of a roadmap created for a virtual camera in a home. The figure on the right shows a probabilistic roadmap tree for docking a spacecraft against a space station (from Jeff Phillips).

  • Topics
    - Principle, rationale, and requirements of probabilistic roadmaps
    - Multi- and Single-query PRM planners
    - Narrow passage issue and expansiveness of free space
    - Probabilistic completeness and convergence of PRM planners
  • Required Readings:
    • Basic paper on probabilistic roadmaps:
      L.E. Kavraki, P. Svestka, J.C. Latombe, and M. Overmars. Probabilistic Roadmaps for Path Planning in High-Dimensional Configuration Spaces. IEEE Transactions on Robotics and Automation, 12(4):566-580, 1996 [pdf]
    • Analysis of probabilistic-roadmap planner:
      D. Hsu, J.C. Latombe, and H. Kurniawati. On the Probabilistic Foundations of Probabilistic Roadmap Planning. Int. J. of Robotics Research, 25(7):627-643, July 2006 [pdf]
  • Other Readings:
    • Computation of neearest neighbors in configuration space:
      Plaku, E. and Kavraki, L. E. . Quantitative Analysis of Nearest Neighbors Search in High-Dimensional Sampling-based Motion Planning. In Workshop on Algorithmic Foundations of Robotics (WAFR), New York City, NY, July 2006. [pdf]
    • Deterministic vs. random sampling:
      S. M. LaValle, M. S. Branicky, and S. R. Lindemann. On the relationship between classical grid search and probabilistic roadmaps. International Journal of Robotics Research, 23(7/8):673-692, July/August 2004.
    • A number of papers on probabilistic roadmaps and related methods can be found at the websites of:
      - Lydia Kavraki
      - Nancy Amato
      - Mark Overmars
      - David Hsu
      - James Kuffner
      - Steve LaValle
      - MOLOG project

  • Slides