DU

Motion Planning Algorithms

Motion Planning Algorithms

Motion Planning involves finding a valid path for a moving object in a given environment. Motion planning algorithms are applied to several disciplines, including robotics, automated design, computational geometry, and biology. Due to their complexity, motion planning problems are generally solved using sampling-based algorithms that rely on an abstraction of the problem called a configuration space (Cspace) and good computational geometry tools.

Motion Planning Algorithms

In my research, I have proposed algorithms that take advantage of workspace guidance to discover important planning regions in the environment efficiently. This has entailed creating a general representation of skeleton annotations to guide traditional planners. I am interested in applying motion planning algorithms to robotics and computational biology applications and studying accessibility.

An illustration of Dynamic Region sampling with PRM. Obstacles are shown in gray. The workspace skeleton is shown in purple. (a) The algorithm samples initial connected components (blue) in regions (green) around each skeleton vertex. (b) Sampling regions expand outward along the skeleton edges. We depict the regions in the location where samples were generated for clarity; in the actual algorithm the regions advance past the newly generated samples. (c) The components in the middle tunnels successfully connect to form bridges, and their regions are released. The outer passages are still expanding.

In my research, I have proposed algorithms that take advantage of workspace guidance to discover important planning regions in the environment efficiently. This has entailed creating a general representation of skeleton annotations to guide traditional planners. I am interested in applying motion planning algorithms to robotics and computational biology applications and studying accessibility.