DOI QR코드

DOI QR Code

Robot Software Framework using Robot Operation System(ROS2) based on Behavior Tree

  • Sangho Lee (Software Engineering Laboratory, Hongik University) ;
  • Hyejin Chang (Dept. of Applied Technology Research, Rastech) ;
  • Seulgi Jeon (Dept. of Applied Technology Research, Rastech) ;
  • Janghwan Kim (Dept. of Computer Engineering, Mokpo National University) ;
  • R. Young Chul, Kim (Dept. Software and Communications Engineering, Hongik University)
  • 투고 : 2023.09.14
  • 심사 : 2023.09.23
  • 발행 : 2023.11.30

초록

As robotic technology expands into various fields, robots need to execute some complicated tasks in diverse environments. However, the previous robotic software solutions were limited to independent systems. We can not adapt to diverse functionalities and environments. This makes it hard to provide rapid and effective services and leads to costs and losses in the development process. To overcome these problems, we propose a robot software framework with behavior trees based on ROS2. This framework simplifies complex robot behaviors through behavior trees and makes it easy to modify, extend, and reuse robot behaviors. Furthermore, ROS2 standardizes connections between software modules, enhances the robot's flexibility, and enables independent development and testing of software. Our framework aims to provide a foundation for high-quality robot service provision by supporting the modularity, reusability, independent development, and testing required by intelligent robots that need to provide services in various environments.

키워드

참고문헌

  1. Kartasev, Mart, "Integrating Reinforcement Learning into Behavior Trees by Hierarchical Composition", MS Thesis, KTH R. Inst. Technol., Stockholm, 2019.
  2. YanChang, Fu, Long, Qin, Quanjun, Yin, "A Reinforcement Learning Behavior Tree Framework for game AI", International Conference on Economics, Social Science, Arts, Education and Management Engineering August 2016. DOI: 10.2991/essaeme-16.2016.120
  3. Hao, Hu, et al. "Self-Adaptive Traffic Control Model with Behavior Trees and Reinforcement Learning for AGV in Industry 4.0", IEEE Transactions on Industrial Informatics December 2021. DOI: 10.1109/TII.2021.3059676
  4. Lei, Li, et al. "Mixed Deep Reinforcement Learning-behavior Tree for Intelligent Agents Design", ICAART February 2021. DOI: 10.5220/0010316901130124
  5. Kartasev Mart, Justin Saler, Petter Orgen "Improving the Performance of Backward Chained Behavior Trees that use Reinforcement Learning", Arxiv Dec 2021. DOI: 10.48550/arXiv.2112.13744
  6. Colledanchise Michele, Petter Ogren "Behavior Trees in Robotics and AI: An Introduction" CRC Press 2018.
  7. Marzinotto Alejandro, et al. "Towards a unified behavior trees framework for robot control", IEEE ICRA MAY 2014. DOI: 10.1109/ICRA.2014.6907656
  8. Jeong Seungwoo, et al. "Behavior Tree-Based Task Planning for Multiple Mobile Robots using a Data Distribution Service", IEEE/ASME International Conference on Advanced Intelligent Mechatronics (AIM) 2022. DOI: 10.1109/AIM52237.2022.9863364
  9. Colledanchise Michele, Lorenzo Natale. "On the implementation of behavior trees in robotics", IEEE Robotics and Automation Letters July 2021. DOI: 10.1109/LRA.2021.3087442
  10. Ghzouli Razan, et al. "Behavior trees in action: a study of robotics applications", Proceedings of the 13th ACM SIGPLAN International Conference on Software Language Engineering November 2020. DOI: 10.1145/3426425.3426942
  11. Colledanchise Michele. "Behavior trees in robotics", Diss. KTH Royal Institute of Technology, 2017.
  12. Paxton Chris, et al. "CoSTAR: Instructing collaborative robots with behavior trees and vision", IEEE ICRA May 2017. DOI: 10.1109/ICRA.2017.7989070