TOPP-RA

Open-source library that takes a path and computes the fastest way to traverse it within limits.

TOPP-RA is a time-optimal path parameterization method proposed by Hung Pham and Quang-Cuong Pham, published in IEEE T-RO (2018) and open-sourced as the Python/C++ library toppra. The problem it solves: a motion planner typically outputs only a sequence of geometric waypoints, with no information about how fast to move at each moment. TOPP-RA discretizes the path into segments and uses reachability analysis (computing, segment by segment, the range of speeds reachable at the next step) to break the problem into a series of small linear programs, producing the shortest-time speed profile subject to constraints on joint velocity, acceleration, and torque. It's more numerically stable and less failure-prone than earlier TOPP methods based on numerical integration. Libraries such as Drake include it built in, and it's commonly chained after path planning to produce an executable trajectory.

ExampleAfter RRT plans a collision-free path for a robot arm, use toppra to redistribute timing according to each joint's velocity and acceleration limits, producing a trajectory ready to send straight to the controller.

Also called

Time-Optimal Path Parameterization based on Reachability Analysis, toppra

Related

Time-Optimal Path Parameterization · Time Parameterization · Trajectory Planning · Ruckig · Drake · MoveIt Motion Planning Framework

Sources

A New Approach to Time-Optimal Path Parameterization based on Reachability Analysis (arXiv)
toppra GitHub