Motion Planning Networks

TL;DR

Motion Planning Networks (MPNet) is a neural network-based algorithm that efficiently generates collision-free paths in high-dimensional environments, outperforming traditional methods in speed and generalization to unseen scenarios.

Contribution

MPNet introduces a novel neural network approach that encodes workspace information from point clouds and produces fast, end-to-end motion plans for complex robotic systems.

Findings

MPNet computes paths in under 1 second across all tested environments.

MPNet generalizes well to unseen environments.

MPNet outperforms existing motion planning algorithms in speed.

Abstract

Fast and efficient motion planning algorithms are crucial for many state-of-the-art robotics applications such as self-driving cars. Existing motion planning methods become ineffective as their computational complexity increases exponentially with the dimensionality of the motion planning problem. To address this issue, we present Motion Planning Networks (MPNet), a neural network-based novel planning algorithm. The proposed method encodes the given workspaces directly from a point cloud measurement and generates the end-to-end collision-free paths for the given start and goal configurations. We evaluate MPNet on various 2D and 3D environments including the planning of a 7 DOF Baxter robot manipulator. The results show that MPNet is not only consistently computationally efficient in all environments but also generalizes to completely unseen environments. The results also show that the computation time of MPNet consistently remains less than 1 second in all presented experiments, which is significantly lower than existing state-of-the-art motion planning algorithms.