Risk Bounded Motion Planning

Robots operating in uncertain environments have to access the risk of collision with dynamic entities present in it. The path planning problem for such robots gets challenging when the distribution governing the stochastic uncertainty is not known exactly rather known to belong to some set of probability distributions satisfying certain conditions. My research is concerned towards creating guaranteed safe paths for robots with stochastic uncertainties. Future research is aimed towards blending the learning techniques to learn both unknown dynamics and uncertainties for the betterment of the safety of these robots.