Autonomous Robot Goal Seeking and Collision Avoidance in the Physical World: An Automated Learning and Evaluation Framework Based on the PPO Method

Mobile robot navigation is a critical aspect of robotics, with applications spanning from service robots to industrial automation. However, navigating in complex and dynamic environments poses many challenges, such as avoiding obstacles, making decisions in real-time, and adapting to new situations....

Full description

Saved in:
Bibliographic Details
Main Authors: Wen-Chung Cheng, Zhen Ni, Xiangnan Zhong, Minghan Wei
Format: Article
Language:English
Published: MDPI AG 2024-11-01
Series:Applied Sciences
Subjects:
Online Access:https://www.mdpi.com/2076-3417/14/23/11020
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850061272584814592
author Wen-Chung Cheng
Zhen Ni
Xiangnan Zhong
Minghan Wei
author_facet Wen-Chung Cheng
Zhen Ni
Xiangnan Zhong
Minghan Wei
author_sort Wen-Chung Cheng
collection DOAJ
description Mobile robot navigation is a critical aspect of robotics, with applications spanning from service robots to industrial automation. However, navigating in complex and dynamic environments poses many challenges, such as avoiding obstacles, making decisions in real-time, and adapting to new situations. Reinforcement Learning (RL) has emerged as a promising approach to enable robots to learn navigation policies from their interactions with the environment. However, application of RL methods to real-world tasks such as mobile robot navigation, and evaluating their performance under various training–testing settings has not been sufficiently researched. In this paper, we have designed an evaluation framework that investigates the RL algorithm’s generalization capability in regard to <i>unseen</i> scenarios in terms of learning convergence and success rates by transferring learned policies in simulation to physical environments. To achieve this, we designed a simulated environment in Gazebo for training the robot over a high number of episodes. The training environment closely mimics the typical indoor scenarios that a mobile robot can encounter, replicating real-world challenges. For evaluation, we designed physical environments with and without unforeseen indoor scenarios. This evaluation framework outputs statistical metrics, which we then use to conduct an extensive study on a deep RL method, namely the proximal policy optimization (PPO). The results provide valuable insights into the strengths and limitations of the method for mobile robot navigation. Our experiments demonstrate that the trained model from simulations can be deployed to the previously <i>unseen</i> physical world with a success rate of over <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mn>88</mn><mo>%</mo></mrow></semantics></math></inline-formula>. The insights gained from our study can assist practitioners and researchers in selecting suitable RL approaches and training–testing settings for their specific robotic navigation tasks.
format Article
id doaj-art-b62bb2207d604e5e963769cda1b8604b
institution DOAJ
issn 2076-3417
language English
publishDate 2024-11-01
publisher MDPI AG
record_format Article
series Applied Sciences
spelling doaj-art-b62bb2207d604e5e963769cda1b8604b2025-08-20T02:50:17ZengMDPI AGApplied Sciences2076-34172024-11-0114231102010.3390/app142311020Autonomous Robot Goal Seeking and Collision Avoidance in the Physical World: An Automated Learning and Evaluation Framework Based on the PPO MethodWen-Chung Cheng0Zhen Ni1Xiangnan Zhong2Minghan Wei3Department of Electrical Engineering and Computer Science, Florida Atlantic University, Boca Raton, FL 33431, USADepartment of Electrical Engineering and Computer Science, Florida Atlantic University, Boca Raton, FL 33431, USADepartment of Electrical Engineering and Computer Science, Florida Atlantic University, Boca Raton, FL 33431, USADepartment of Electrical Engineering and Computer Science, Florida Atlantic University, Boca Raton, FL 33431, USAMobile robot navigation is a critical aspect of robotics, with applications spanning from service robots to industrial automation. However, navigating in complex and dynamic environments poses many challenges, such as avoiding obstacles, making decisions in real-time, and adapting to new situations. Reinforcement Learning (RL) has emerged as a promising approach to enable robots to learn navigation policies from their interactions with the environment. However, application of RL methods to real-world tasks such as mobile robot navigation, and evaluating their performance under various training–testing settings has not been sufficiently researched. In this paper, we have designed an evaluation framework that investigates the RL algorithm’s generalization capability in regard to <i>unseen</i> scenarios in terms of learning convergence and success rates by transferring learned policies in simulation to physical environments. To achieve this, we designed a simulated environment in Gazebo for training the robot over a high number of episodes. The training environment closely mimics the typical indoor scenarios that a mobile robot can encounter, replicating real-world challenges. For evaluation, we designed physical environments with and without unforeseen indoor scenarios. This evaluation framework outputs statistical metrics, which we then use to conduct an extensive study on a deep RL method, namely the proximal policy optimization (PPO). The results provide valuable insights into the strengths and limitations of the method for mobile robot navigation. Our experiments demonstrate that the trained model from simulations can be deployed to the previously <i>unseen</i> physical world with a success rate of over <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mn>88</mn><mo>%</mo></mrow></semantics></math></inline-formula>. The insights gained from our study can assist practitioners and researchers in selecting suitable RL approaches and training–testing settings for their specific robotic navigation tasks.https://www.mdpi.com/2076-3417/14/23/11020automated learning frameworkplatform developmentautonomous robotdeep reinforcement learningphysical implementationcollision avoidance
spellingShingle Wen-Chung Cheng
Zhen Ni
Xiangnan Zhong
Minghan Wei
Autonomous Robot Goal Seeking and Collision Avoidance in the Physical World: An Automated Learning and Evaluation Framework Based on the PPO Method
Applied Sciences
automated learning framework
platform development
autonomous robot
deep reinforcement learning
physical implementation
collision avoidance
title Autonomous Robot Goal Seeking and Collision Avoidance in the Physical World: An Automated Learning and Evaluation Framework Based on the PPO Method
title_full Autonomous Robot Goal Seeking and Collision Avoidance in the Physical World: An Automated Learning and Evaluation Framework Based on the PPO Method
title_fullStr Autonomous Robot Goal Seeking and Collision Avoidance in the Physical World: An Automated Learning and Evaluation Framework Based on the PPO Method
title_full_unstemmed Autonomous Robot Goal Seeking and Collision Avoidance in the Physical World: An Automated Learning and Evaluation Framework Based on the PPO Method
title_short Autonomous Robot Goal Seeking and Collision Avoidance in the Physical World: An Automated Learning and Evaluation Framework Based on the PPO Method
title_sort autonomous robot goal seeking and collision avoidance in the physical world an automated learning and evaluation framework based on the ppo method
topic automated learning framework
platform development
autonomous robot
deep reinforcement learning
physical implementation
collision avoidance
url https://www.mdpi.com/2076-3417/14/23/11020
work_keys_str_mv AT wenchungcheng autonomousrobotgoalseekingandcollisionavoidanceinthephysicalworldanautomatedlearningandevaluationframeworkbasedontheppomethod
AT zhenni autonomousrobotgoalseekingandcollisionavoidanceinthephysicalworldanautomatedlearningandevaluationframeworkbasedontheppomethod
AT xiangnanzhong autonomousrobotgoalseekingandcollisionavoidanceinthephysicalworldanautomatedlearningandevaluationframeworkbasedontheppomethod
AT minghanwei autonomousrobotgoalseekingandcollisionavoidanceinthephysicalworldanautomatedlearningandevaluationframeworkbasedontheppomethod