Deep reinforcement learning for map-less goal-driven robot navigation

Dobrevski, Matej (author), Skočaj, Danijel (author)

Mobile robots that operate in real-world environments need to be able to safely navigate their surroundings. Obstacle avoidance and path planning are crucial capabilities for achieving autonomy of such systems. However, for new or dynamic environments, navigation methods that rely on an explicit map of the environment can be impractical or even impossible to use. We present a new local navigation method for steering the robot to global goals without relying on an explicit map of the environment. The proposed navigation model is trained in a deep reinforcement learning framework based on Advantage Actor–Critic method and is able to directly translate robot observations to movement commands. We evaluate and compare the proposed navigation method with standard map-based approaches on several navigation scenarios in simulation and demonstrate that our method is able to navigate the robot also without the map or when the map gets corrupted, while the standard approaches fail. We also show that our method can be directly transferred to a real robot.

• Language: English
• Keywords: mobile robotics, reinforcement learning, navigation, deep learning
• Work type: Article
• Typology: 1.01 - Original Scientific Article
• Organization: FRI - Faculty of Computer and Information Science
• Publication status: Published
• Publication version: Version of Record
• Year: 2021
• Number of pages: 13 str.
• Numbering: Vol. 18, iss. 1
• PID: 20.500.12556/RUL-144122
• UDC: 007.52:004.8
• ISSN on article: 1729-8814
• DOI: 10.1177/1729881421992621
• COBISS.SI-ID: 60005379
• Publication date in RUL: 01.02.2023

Record is a part of a journal

• Title: International journal of advanced robotic systems
• Shortened title: Int. j. adv. robot. syst.
• Publisher: SAGE
• ISSN: 1729-8814
• COBISS.SI-ID: 9061716

Licences

• License: CC BY 4.0, Creative Commons Attribution 4.0 International
• Link: http://creativecommons.org/licenses/by/4.0/
• Description: This is the standard Creative Commons license that gives others maximum freedom to do what they want with the work as long as they credit the author.

Secondary language

• Language: Slovenian
• Keywords: mobilna robotika, spodbujevalno učenje, navigacija, globoko učenje

Projects

• Funder: ARRS - Slovenian Research Agency
• Project number: J2-9433
• Name: Iskanje nekonsistentnosti v kompleksnih slikovnih podatkih z globokim učenjem

• Funder: ARRS - Slovenian Research Agency
• Project number: P2-0214
• Name: Računalniški vid