Kavraki Lab

Physical and Biological Computing

• Motion Planning with Kinodynamic and Physics-based Constraints

  • Kinodynamic Planning by Interior-Exterior Cell Exploration (KPIECE)
  • Path-Directed Subdivision Tree Exploration (PDST)
  • Synergistic Combination of Layers of Planning (SyCLoP)
  • Guided Expansive Spaces Trees

• Planning for Systems with Complex Dynamics from High-Level Specifications

  • Temporal Logic Motion Planning

• Hybrid Systems

  • Overview
  • Counterexample Search-based Approach for Safety Analysis of Hybrid Systems (HyDICE)
  • Falsification of LTL Safety Specifications using Sampling-based Algorithms (TemporalHyDICE)

• Replanning

  • Replanning: A Motion Planning Strategy for Solving Hard Kinodynamic Problems
  • Distributed and Safe Replanning for Communicating Vehicles
  • Greedy but Safe Replanning for a Single Vehicle with Dynamics

• Mobile Robotics

  • Bearing-Only Navigation Without a Map or Localization
  • Bearing-Only Simultaneous Localization and Mapping
  • Localization Using the Signal Strength of Wireless Ethernet Network Packets

• Earlier work on Path Planning

  • Probabilistic Roadmap Method (PRM) and its Analysis
  • Lazy PRM - Medial Axis PRM - PRM for Manipulation
  • Deformable Parts - Simulated Knot Tying - Mathematical Knot Untangling
  • Planning for Inspection
  • Decomposition-based Motion Planning
  • Distributed Sampling-based Roadmap of Trees (DSRT)
  • Sampling-based Roadmap of Trees (SRT)

• Proximity Relations in High-Dimensional Metric Spaces

  • A High-Performance Distributed Platform For the Nearest-Neighbors Graph (DKNNG)
  • Approximate Nearest Neighbors for Motion Planning and Nonlinear Dimensionality Reduction (DPES and hcDPES)

• Automation and Manipulation

  • Part Orientation Using Microelectromechanical (MEMS) arrays

Teaching and Educational Material

COMP450: Algorithmic Robotics is taught regularly at Rice University. The material of the course is under Rice's OWLSPACE.

We have also organized workshops and tutorials related to some of the above topics:

• At IROS 2011 conference in San Francisco, we organized a full date tutorial "Motion Planning for Physical Robots" which is an extensive presentation of the OMPL library.
• At the IROS 2008 conference in Nice, France, we organized a tutorial titled "Motion planning with the OOPSMP toolkit: A hands-on tutorial on using state-of-the-art motion planning algorithms."
• The workshop on Algorithmic Motion Planning for Autonomous Robots in Challenging Environments took place in conjunction with the IROS 2007 conference, held in San Diego, CA between Oct. 29 to Nov. 2.