Motion Planning Among Dynamic, Decision-Making Agents with Deep Reinforcement Learning

TL;DR

This paper presents a deep reinforcement learning algorithm for robot navigation among dynamic agents that adapts to varying numbers of agents and does not assume specific agent behaviors, improving safety and efficiency.

Contribution

It introduces a novel LSTM-based strategy allowing the algorithm to observe an arbitrary number of agents without fixed input size, and extends previous methods to more realistic multi-agent scenarios.

Findings

Outperforms previous approaches as the number of agents increases

Successfully navigates a robotic vehicle at human walking speed without 3D Lidar

Demonstrates robustness in simulation and real-world tests

Abstract

Robots that navigate among pedestrians use collision avoidance algorithms to enable safe and efficient operation. Recent works present deep reinforcement learning as a framework to model the complex interactions and cooperation. However, they are implemented using key assumptions about other agents' behavior that deviate from reality as the number of agents in the environment increases. This work extends our previous approach to develop an algorithm that learns collision avoidance among a variety of types of dynamic agents without assuming they follow any particular behavior rules. This work also introduces a strategy using LSTM that enables the algorithm to use observations of an arbitrary number of other agents, instead of previous methods that have a fixed observation size. The proposed algorithm outperforms our previous approach in simulation as the number of agents increases, and the algorithm is demonstrated on a fully autonomous robotic vehicle traveling at human walking speed, without the use of a 3D Lidar.