RRT: Real-Time Motion Planning/Replanning for Environments with Unpredictable Obstacles

Dynamic environments have obstacles that unpredictably appear, disappear, or move. We present the first sampling-based motion planning algorithm for real-time navigation through dynamic environments. Our algorithm, RRT, refines and repairs the same search-graph over the entire duration of navigation, i.e., despite changing obstacles and/or robot location. Whenever such changes are observed, e.g., by onboard sensors, a graph rewiring cascade quickly updates the search-graph and repairs its shortest-path-to-goal sub-tree. Both graph and tree are built directly in the robot’s state space; thus, the resulting plan(s) respect the kinematics of the robot and continue to improve during navigation. RRT is probabilistically complete and makes no distinction between local and global planning, yet it reacts quickly enough for real-time highspeed navigation though unpredictably changing environments. Low information transfer time is essential for enabling RRT to react quickly in dynamic environments; we prove that the information transfer time required to inform a graph of size n about an -cost decrease is O (n logn) for RRT—faster than RRT* Ω ( n( n logn ) ) and RRT ω ( n log n ) . RRT is also competitive in static environments—where it has the same amortized per iteration runtime as RRT and RRT* Θ (logn) and is faster than RRT ω ( log n ) . In order to achieve O (logn) iteration time, each node maintains a set of O (logn) expected neighbors, and the search graph maintains -consistency for a predefined . The utility of RRT in dynamic environments is demonstrated by real-time simulations. Experiments in static environments confirm its properties vs. RRT* and RRT.