Architecture
Navigating a nonholonomic robot in a cluttered, unknown environment requires accurate perception and precise motion control for real-time collision avoidance. This paper presents NeuPAN: a real-time, highly accurate, map-free, easy-to-deploy, and environment-invariant robot motion planner. Leveraging a tightly coupled perception-to-control framework, NeuPAN has two key innovations compared to existing approaches:
1) it directly maps raw point cloud data to a latent distance feature space for collision-free motion generation, avoiding error propagation from the perception to control pipeline;
2) it is interpretable from an end-to-end model-based learning perspective.
The crux of NeuPAN is solving an end-to-end mathematical model with numerous point-level constraints using a plug-and-play (PnP) proximal alternating-minimization network (PAN), incorporating neurons in the loop. This allows NeuPAN to generate real-time, physically interpretable motions. It seamlessly integrates data and knowledge engines, and its network parameters can be fine-tuned via backpropagation. We evaluate NeuPAN on a ground mobile robot, a wheel-legged robot, and an autonomous vehicle, in extensive simulated and real-world environments. Results demonstrate that NeuPAN outperforms existing baselines in terms of accuracy, efficiency, robustness, and generalization capabilities across various environments, including the cluttered sandbox, office, corridor, and parking lot. We show that NeuPAN works well in unknown and unstructured environments with arbitrarily shaped objects, transforming impassable paths into passable ones.
1. High Accuracy Control: Compared with existing planners (e.g., TEB), NeuPAN achieves a 2X improvement in accuracy, enabling the robot to navigate extremely narrow paths that others cannot.
2. Real Time Implementation: Compared with other optimization-based planners, NeuPAN is faster, enabling real-time implementation on real-world platforms.
3. Easy to Deploy: Compared to other end-to-end learning-based navigation policies, the training process of NeuPAN (DNUE) takes only approximately one hour to complete. Additionally, for a specific robot with a fixed shape, the trained model can be used anywhere without the need for retraining.
4. Generalization and Adaptability: NeuPAN is an MPC-based planner that can adapt to different robot platforms and environments by adjusting its parameters and constraints simply.
@article{han2024neupan,
title = {NeuPAN: Direct Point Robot Navigation with End-to-End Model-based Learning},
author = {Han, Ruihua and Wang, Shuai and Wang, Shuaijun and Zhang, Zeqing and Chen, Jianjun and Lin, Shijie and Li, Chengyang and Xu, Chengzhong and Eldar, Yonina C and Hao, Qi and others},
journal = {arXiv preprint arXiv:2403.06828},
year = {2024},
}We thank the authors of Nerfies and Academic Project Page Template that kindly open sourced the website template.