Robust Motion Planning

Traditional motion planning methods often fail to provide formal robust safety and performance guarantees. My research leverages Hamilton-Jacobi reachability analysis and differential games to compute provably safe motion plans in time-varying and adversarial environments under parametric uncertainty. This work also addresses stochastic trajectory optimization for handling uncertainties in contact interactions.

Related Publications

    2. Gazar22 icon
      []
      A. Gazar, M. Khadiv, Sébastien Kleff, A. D. Prete, L. Righetti, "Nonlinear stochastic trajectory optimization for centroidal momentum motion generation of legged robots", The International Symposium of Robotics Research, ISR 2022, 2022.
      Paper Video
    4. KleffRobotica20 icon
      []
      Sébastien Kleff, N. Li, "Robust Motion Planning in Dynamic Environments Based on Sampled-Data Hamilton-Jacobi Reachability", Robotica, 2020.
      Paper
    6. KleffCCDC2018 icon
      []
      Sébastien Kleff, N. Li, "A sampled-data Hamilton-Jacobi reachability approach to safe and robust motion planning", Chinese Control And Decision Conference (CCDC), 2018.
      Paper