Planning Under Uncertainty

Generating robust plans in uncertain scenarios.

Planning under uncertainty is a significant challenge in robotics, as robots often rely on incomplete geometric information due to sensing limitations. For instance, a robot might only have a partial view of an object, leading to collision with unseen parts. To address this, the ELPIS lab is conducting research in effective motion planning under uncertainty.

Relevant Publications

  1. RSS
    Meta-Policy Learning over Plan Ensembles for Robust Articulated Object Manipulation
    C. Chamzas, C. Garrett, B. Sundaralingam, L. Kavraki, and D. Fox
    In RSS 2023: Workshop on Learning for Task and Motion Planning, 2023
  2. Human-Guided Motion Planning in Partially Observable Environments
    C. Quintero-Peña*, C. Chamzas*, Z. Sun, V. Unhelkar, and L. E. Kavraki
    In IEEE International Conference on Robotics and Automation, 2022
  3. RAL
    MotionBenchMaker: A Tool to Generate and Benchmark Motion Planning Datasets
    C. Chamzas, C. Quintero-Peña, Z. Kingston, A. Orthey, D. Rakita, M. Gleicher, M. Toussaint, and L. E. Kavraki
    IEEE Robotics and Automation Letters, 2022
  4. Motion Planning via Bayesian Learning in the Dark
    C. Quintero-Peña*, C. Chamzas*, V. Unhelkar, and L. E. Kavraki
    In ICRA 2021: Workshop on Machine Learning for Motion Planning, 2021
    (Spotlight)