StratMamba: Strategic and Reactive Stream Partitioning for Path-Efficient LiDAR-Based Obstacle Avoidance

2026-07-14

AI Summary

This work addresses the challenge of simultaneously achieving efficient real-time obstacle avoidance and long-horizon path planning for robots operating in complex, cluttered environments. The authors propose StratMamba, a novel dual-stream Mamba-based architecture that decouples perception and planning into two complementary temporal streams: a high-frequency LiDAR stream employing a fast-decaying memory mechanism for immediate collision avoidance, and a low-frequency goal stream utilizing a slow-decaying memory mechanism to support global strategic planning. Integrating temporal modeling, dual-stream memory dynamics, and reinforcement learning, the method is deployed and evaluated in IsaacLab and Gazebo simulations as well as on a Unitree GO1 quadrupedal robot. Experiments demonstrate state-of-the-art performance in simulation, achieving the lowest timeout rate, fastest navigation speed (median 576 steps), and highest path efficiency (0.915), while real-world tests confirm robust operation under long-range LiDAR sensing conditions.

Abstract

This paper proposes StratMamba, a dual-stream Mamba-based temporal modeling architecture, to more efficiently capture long-horizon temporal dependencies required for robot navigation in complex and obstacle-rich environments. StratMamba leverages a combination of fast-decay and slow-decay memory architectures, where the fast-decay component processes high-frequency LiDAR data for reactive obstacle avoidance, while the slow-decay component maintains longer-horizon goal information for strategic planning. We perform extensive evaluations of different obstacle avoidance scenarios in IsaacLab and Gazebo, while also validating successful sim-to-real deployment on a Unitree GO1 quadruped robot navigating in the presence of static/dynamic obstacles. Comparisons with other temporal RL baselines, such as LSTM, Transformer, and Vanilla-Mamba, show that our StratMamba achieves exceptional temporal reasoning efficiency with a lower timeout rate, while maintaining the fastest navigation speed (576 median steps, 5.0% better than Vanilla-Mamba). It also achieves the highest path optimality (0.915 path efficiency) across all baselines. Real-world evaluation reveals that StratMamba maintains more robust performance across extended LiDAR ranges compared to vanilla Mamba and the Transformer, demonstrating that dual-stream partitioning effectively balances reactive safety with strategic navigation under challenging sensing conditions.

Problem

• obstacle avoidance
• temporal modeling
• LiDAR-based navigation
• strategic planning
• reactive control

Innovation

• StratMamba
• dual-stream partitioning
• temporal modeling
• LiDAR-based navigation
• reactive-strategic balance