Learning World Transition Model for Socially Aware Robot Navigation

TL;DR

This paper introduces a model-based reinforcement learning approach for socially aware robot navigation in crowded environments, utilizing a deep transition model to predict dynamic surroundings and reduce real data requirements.

Contribution

It proposes a novel transition model that predicts environment dynamics from laser scans, enabling efficient training with less real interaction data for social navigation.

Findings

Achieves high success rate in social navigation with less real data

Effective in both simulation and real-world scenarios

Guides robots to targets while avoiding pedestrians socially

Abstract

Moving in dynamic pedestrian environments is one of the important requirements for autonomous mobile robots. We present a model-based reinforcement learning approach for robots to navigate through crowded environments. The navigation policy is trained with both real interaction data from multi-agent simulation and virtual data from a deep transition model that predicts the evolution of surrounding dynamics of mobile robots. The model takes laser scan sequence and robot's own state as input and outputs steering control. The laser sequence is further transformed into stacked local obstacle maps disentangled from robot's ego motion to separate the static and dynamic obstacles, simplifying the model training. We observe that our method can be trained with significantly less real interaction data in simulator but achieve similar level of success rate in social navigation task compared with other methods. Experiments were conducted in multiple social scenarios both in simulation and on real robots, the learned policy can guide the robots to the final targets successfully while avoiding pedestrians in a socially compliant manner. Code is available at https://github.com/YuxiangCui/model-based-social-navigation