LSTP-Nav: Lightweight Spatiotemporal Policy for Map-free Multi-agent Navigation with LiDAR

Code：https://github.com/xrdiao/LSTP-Nav

Safe map-free multi-robot navigation requires robots to make real-time decisions from partial and noisy local observations in dynamic, unstructured environments. Existing approaches often depend on prior maps, computationally intensive perception pipelines, or carefully tuned interaction models, which limit their robustness on resource-constrained platforms. This paper proposes LSTP-Nav, a lightweight, decentralized navigation framework built on LSTP-Net that maps stacked 2D LiDAR observations, goal information, and velocity feedback directly to action. LSTP-Net combines GRU-based temporal modeling with attention to extract interaction-critical spatiotemporal cues at low computational cost. We further introduce an HS reward to provide smooth, heading-aware safety feedback, and develop PhysReplay-SimLab to improve training effectiveness through local replay of near-failure interactions. Extensive simulation and real-world experiments show that LSTP-Nav outperforms representative baselines, achieving 98.6%–100.0% success in single-robot settings and 97.8%–99.0% success in 10-robot settings with 5–35 obstacles. The policy also remains robust to LiDAR degradation, nonconvex layouts, random seeds, and large-scale density tests, while achieving above 40 Hz onboard inference on a Raspberry Pi 3 Model B with zero-shot sim-to-real transfer. Overall, LSTP-Nav provides a practical, deployable solution for real-time, map-free multi-agent navigation on computationally constrained mobile robots.

We demonstrate the performance of the LSTP-Nav framework, with the agent and obstacles randomly placed in a 10 m × 10 m area 

The green robot employs the LSTP-Net policy, while the yellow agent (controlled by a human operator) serves as the dynamic goal. The other red agents are deployed with a random walk policy, ORCA, Linear policy, GRU policy, CNN policy [25], or  LSTP-Net policy.

We compare the HybridRL-CA [23] and conduct ablation studies (Linear and GRU policies) against the LSTP-Net policy in both single-agent and multi-agent scenarios. 

[21] T. Fan, P. Long, W. Liu, and J. Pan, “Distributed multi-robot collision avoidance via deep reinforcement learning for navigation in complex scenarios,” The International Journal of Robotics Research, 2020.