Optimal-Horizon Social Robot Navigation in Heterogeneous Crowds

2026-02-28

AI Summary

This work addresses the “freezing robot” problem in social navigation among heterogeneous crowds, which arises when model predictive control (MPC) with a fixed prediction horizon fails to adapt to dynamic social contexts. The authors propose the first approach that treats the MPC prediction horizon as an optimizable variable driven by social context. By leveraging a spatiotemporal Transformer to infer pedestrian cooperativeness from local trajectories and integrating reinforcement learning to dynamically adjust the planning horizon online, the method jointly optimizes task objectives and socially aware safety constraints. Experimental results demonstrate that the proposed framework improves navigation success rate by 6.8%, reduces collisions by 50%, shortens navigation time by 19%, and achieves a remarkably low timeout rate of only 0.8% compared to state-of-the-art baselines.

Technology Category

• Humans and AI: Human-Aware Planning and Behavior Prediction
• Planning, Routing, and Scheduling: Replanning and Plan Repair
• Natural Language Processing: Safety and Robustness

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Navigating social robots in dense, dynamic crowds is challenging due to environmental uncertainty and complex human-robot interactions. While Model Predictive Control (MPC) offers strong real-time performance, its reliance on a fixed prediction horizon limits adaptability to changing environments and social dynamics. Furthermore, most MPC approaches treat pedestrians as homogeneous obstacles, ignoring social heterogeneity and cooperative or adversarial interactions, which often causes the Frozen Robot Problem in partially observable real-world environments. In this paper, we identify the planning horizon as a socially conditioned decision variable rather than a fixed design choice. Building on this insight, we propose an optimal-horizon social navigation framework that optimizes MPC foresight online according to inferred social context. A spatio-temporal Transformer infers pedestrian cooperation attributes from local trajectory observations, which serve as social priors for a reinforcement learning policy that optimally selects the prediction horizon under a task-driven objective. The resulting horizon-aware MPC incorporates socially conditioned safety constraints to balance navigation efficiency and interaction safety. Extensive simulations and real-world robot experiments demonstrate that optimal foresight selection is critical for robust social navigation in partially observable crowds. Compared to state-of-the-art baselines, the proposed approach achieves a 6.8\% improvement in success rate, reduces collisions by 50\%, and shortens navigation time by 19\%, with a low timeout rate of 0.8\%, validating the necessity of socially optimal planning horizons for efficient and safe robot navigation in crowded environments. Code and videos are available at Under Review.

Problem

• social robot navigation
• heterogeneous crowds
• prediction horizon
• frozen robot problem
• human-robot interaction

Innovation

• optimal-horizon MPC
• socially-aware navigation
• spatio-temporal Transformer
• heterogeneous crowds
• reinforcement learning for horizon selection