Realtime Informed Path Sampling for Motion Planning Search

Abstract

Mobile robot motions often originate from an uninformed path sampling process such as random or low-dispersion sampling. We demonstrate an alternative approach to path sampling that closes the loop on the expensive collision-testing process. Although all necessary information for collision-testing a path is known to the planner, that information is typically stored in a relatively unavailable form in a costmap or obstacle map. By summarizing the most salient data in a more accessible form, our process delivers a denser sampling of the free path space per unit time than open-loop sampling techniques. We obtain this result by probabilistically modeling-in real time and with minimal information-the locations of obstacles and free space, based on collision-test results. We present CALM, the combined adaptive locality model, along with an algorithm to bias path sampling based on the model's predictions. We provide experimental results in simulation for motion planning on mobile robots, demonstrating up to a 330% increase in paths surviving collision test.

Autonomous Waypoint Generation with Safety Guarantees: On-Line Motion Planning in Unknown Environments

“…Kuwata et al [7] presented non-fixed random path-sets using RRT. A similar work is fixed path-set, presented by Knepper et al [5] (and references therein). Path-set methods rely on a global guidance function to guide the search towards the goal, and often constructing such a function requires map of the environment.…”

Conventional, Heuristic and Learning-Based Robot Motion Planning: Reviewing Frameworks of Current Practical Significance

“…Thus, the information required for collision testing needs to be accessible to the planner. However, this might not be feasible in many scenarios [44].…”

“…When it comes to computations, the cost functions related to collision testing can be enhanced by summarizing the most salient data through probabilistic modeling [44]. Directly searching a dense motion planning RM is run in O(V log V + E) ≈ O n 2 time, where n is the number of vertices.…”