Open Motion Planning Library (OMPL)

An open-source motion planning library from Rice University that bundles dozens of sampling-based planning algorithms.

OMPL is an open-source C++ motion planning library, with Python bindings, developed by the Kavraki Lab at Rice University. It implements dozens of sampling-based planning algorithms, including RRT, RRT-Connect, RRT*, PRM, and BIT*. Its only job is to find a collision-free path through configuration space — it doesn't handle collision checking or robot modeling itself, relying instead on an externally supplied function that answers whether a given state is valid. That decoupling is exactly why MoveIt uses it as its default planning backend, so a lot of collision-avoidance planning for robot arms ultimately runs on OMPL underneath. It's also a convenient way to compare different planning algorithms when learning about motion planning.

ExampleWhen MoveIt plans a path for a UR5e to avoid a cup on a table, it calls OMPL's RRTConnect by default.

Also called

OMPL

Related

MoveIt Motion Planning Framework · Motion Planning · Sampling-Based Planning · Rapidly-exploring Random Tree · Probabilistic Roadmap · Flexible Collision Library (FCL)

Sources

The Open Motion Planning Library