Following the Human Thread in Social Navigation

2024-04-17

arXiv.org

AI Summary

Social navigation in human-robot coexistence scenarios faces the core challenge of real-time human motion intention perception, requiring robots to proactively yield while maintaining collaborative distances and avoiding collisions—complicated by partial observability and modeling complexity of human trajectories in egocentric views. To address this, we propose SDA, the first method that infers social dynamics from state-action history. Our approach employs a two-stage reinforcement learning framework enabling generalization from trajectory-supervised learning to purely history-driven navigation. Integrating trajectory encoding, implicit dynamic reasoning, and Habitat 3.0 simulation, SDA predicts and responds to human intentions with high accuracy—without requiring real-time human trajectory inputs. Evaluated on human-following tasks, SDA achieves state-of-the-art performance, significantly improving both runtime efficiency and robustness against occlusion, sensor noise, and dynamic environment changes.

Technology Category

• Humans and AI: Human-Aware Planning and Behavior Prediction
• Intelligent Robots: State Estimation
• Planning, Routing, and Scheduling: Replanning and Plan Repair

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

The success of collaboration between humans and robots in shared environments relies on the robot's real-time adaptation to human motion. Specifically, in Social Navigation, the agent should be close enough to assist but ready to back up to let the human move freely, avoiding collisions. Human trajectories emerge as crucial cues in Social Navigation, but they are partially observable from the robot's egocentric view and computationally complex to process. We present the first Social Dynamics Adaptation model (SDA) based on the robot's state-action history to infer the social dynamics. We propose a two-stage Reinforcement Learning framework: the first learns to encode the human trajectories into social dynamics and learns a motion policy conditioned on this encoded information, the current status, and the previous action. Here, the trajectories are fully visible, i.e., assumed as privileged information. In the second stage, the trained policy operates without direct access to trajectories. Instead, the model infers the social dynamics solely from the history of previous actions and statuses in real-time. Tested on the novel Habitat 3.0 platform, SDA sets a novel state-of-the-art (SotA) performance in finding and following humans. The code can be found at https://github.com/L-Scofano/SDA.

Problem

• real-time adaptation to human motion
• infer social dynamics from history
• avoid collisions while assisting humans

Innovation

• Reinforcement Learning framework
• Social Dynamics Adaptation model
• real-time human trajectory inference