Steven M. LaValle

Software


Bang-Bang Steering

This open-source Python code performs time-optimal steering for a vector of double integrators. We used it to make RRT-based kinodynamic planning three orders of magnitude faster. It is a simple module that can be used in other planning or learning algorithms. It was written by Alexander LaValle. See the related IEEE IROS 2023 paper by A. J. LaValle. B. Sakcak, and S. M. LaValle.

Motion Strategy Library

This is the first open-source general-purpose motion planning library. It was developed in 2000 for implementing and comparing motion planning algorithms, for use in research, education, and industry. See the Motion Strategy Library page.

MPNN: A Nearest-Neighbor Library for Motion Planning

A C++ library, written by Anna Yershova, that uses Kd-trees adapted to topological spaces that arise in motion planning. This enables fast nearest-neighbor computations in sampling-based motion planning algorithms.

Sampling the Space of 3D Rotations, SO(3)

A C++ library that generates sequences of samples that are close to uniform and have regular neighborhood structure.

Embarrassingly Simple RRT Code

This Python code uses the PyGame package, which is easy to install (in Ubuntu, it is available by default from the package installer). Note: This only draws the edges, rather than maintain a data structure.

Planning Optimal Paths for Multiple Agents on Graphs

Written by Jingjin Yu. The Java code is available here and the associated paper is here.