REVIEW 4 cited by
Neural MP: A Generalist Neural Motion Planner
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
read the original abstract
The current paradigm for motion planning generates solutions from scratch for every new problem, which consumes significant amounts of time and computational resources. For complex, cluttered scenes, motion planning approaches can often take minutes to produce a solution, while humans are able to accurately and safely reach any goal in seconds by leveraging their prior experience. We seek to do the same by applying data-driven learning at scale to the problem of motion planning. Our approach builds a large number of complex scenes in simulation, collects expert data from a motion planner, then distills it into a reactive generalist policy. We then combine this with lightweight optimization to obtain a safe path for real world deployment. We perform a thorough evaluation of our method on 64 motion planning tasks across four diverse environments with randomized poses, scenes and obstacles, in the real world, demonstrating an improvement of 23%, 17% and 79% motion planning success rate over state of the art sampling, optimization and learning based planning methods. Video results available at mihdalal.github.io/neuralmotionplanner
Forward citations
Cited by 4 Pith papers
-
SplatCtrl: Perception-Action Coupling via Gaussian Scene Representations and Reactive Robot Control
SplatCtrl couples real-time isotropic Gaussian scene reconstruction from RGB-D with continuous GPDF-derived SDFs inside control-barrier QP-IK for collision-free 6-DoF robot motion in dynamic environments.
-
Deep Reactive Policy: Learning Reactive Manipulator Motion Planning for Dynamic Environments
DRP couples a transformer policy pretrained on 10M cuRobo trajectories and refined by student-teacher finetuning with a point-cloud reactive goal proposer (DCP-RMP) to outperform prior planners in dynamic manipulation.
-
Growing Trees with an Agent: Accelerating RRTs with Learned, Multi-Step Episodic Exploration
A reinforcement learning policy that emits multi-step exploration episodes can replace random sampling in RRT-style planners, yielding reported order-of-magnitude speedups and higher success rates in simulated navigation.
-
SAMP: Spatial Anchor-based Motion Policy for Collision-Aware Robotic Manipulators
SAMP aligns environment and robot signed distance fields on a shared spatial grid to train a collision-aware neural motion policy for manipulators.
Discussion (0). Sign in to comment.