Robust Reinforcement Learning-based Autonomous Driving Agent for Simulation and Real World

TL;DR

This paper introduces a Deep Reinforcement Learning algorithm using Deep Q-Networks for autonomous driving that is trained in simulation and successfully transferred to real-world environments, demonstrating effective lane following.

Contribution

The paper presents a novel DRL-based method that enables autonomous driving agents trained in simulation to perform effectively in real-world settings.

Findings

Agent trained in simulation performs well in real-world environment

Method requires limited hardware resources

Performance is comparable to state-of-the-art approaches

Abstract

Deep Reinforcement Learning (DRL) has been successfully used to solve different challenges, e.g. complex board and computer games, recently. However, solving real-world robotics tasks with DRL seems to be a more difficult challenge. The desired approach would be to train the agent in a simulator and transfer it to the real world. Still, models trained in a simulator tend to perform poorly in real-world environments due to the differences. In this paper, we present a DRL-based algorithm that is capable of performing autonomous robot control using Deep Q-Networks (DQN). In our approach, the agent is trained in a simulated environment and it is able to navigate both in a simulated and real-world environment. The method is evaluated in the Duckietown environment, where the agent has to follow the lane based on a monocular camera input. The trained agent is able to run on limited hardware resources and its performance is comparable to state-of-the-art approaches.