Motion planning accelerator speeds up robot path finding by up to 44 times
ScaleMPA: Rethinking Scalable RRT* Acceleration With a Grid-Native Representation
RoboticsHardware Architecture
Summary
Finding the best path for a robot to move in complex spaces is hard and slow. Previous methods organized the search as a tree, which limited speed and parallel processing. The authors created ScaleMPA, which uses a grid-based system instead, allowing faster and more parallel searches. This new method also uses clever tricks like multi-resolution grids and hashing to handle big spaces efficiently. As a result, ScaleMPA runs much faster than existing systems, enabling real-time planning in milliseconds.
What this means in practice
- •For robotics engineers: Accelerate robot navigation by reducing motion planning delays in complex environments with ScaleMPA’s grid-based planning hardware.
- •For autonomous vehicle developers: Improve real-time path planning for autonomous cars by integrating ScaleMPA’s fast grid-native planning accelerator into vehicle control systems.
Authors
Zilong Wang, Yuzhou Chen, Xinyue He, Chen Zhang, Guanghui He
Abstract
Real-time motion planning remains challenging in large and high-dimensional environments. Prior acceleration of RRT* follows tree-centric state organization, which reduces per-query cost but preserves superlinear end-to-end complexity and limits parallelism through structural dependencies. This paper presents ScaleMPA, a motion-planning accelerator that rethinks RRT* with a grid-native representation. By replacing hierarchical traversal with direct grid-based access, ScaleMPA reduces the planner critical path and exposes fine-grained parallelism. To make this reformulation practical under sparse high-dimensional planning, ScaleMPA further proposes a multi-resolution grid search engine and a hash-grid memory system. Implemented in 28 nm CMOS, ScaleMPA achieves millisecond-level planning latency and delivers 4.7$\times$--44.4$\times$ speedup over state-of-the-art motion-planning accelerators.