Design and Preliminary Results of Novel Situational Awareness Systemfor Autonomous Ship based on Artificial Intelligence Techniques인공지능 기술에 기반한 자율운항선박의 새로운 상황인식 시스템의 설계와 초기 결과

Cited 0 time in webofscience Cited 0 time in scopus
  • Hit : 17
  • Download : 0
With recent advancements in artificial intelligence technologies, the academia and industry have heightened performance expectations when designing autonomous ships, and major ship building companies and governments have been conducting large-scale research projects around the world. This paper proposes a novel design concept, and presents the key features and preliminary results of a situational awareness system for autonomous ships, named the iSAS(Intelligent Situational Awareness System), and is being developed as part of the Korea Autonomous Surface Ships (KASS) project launched in April 2020. The iSAS comprises deep-learning algorithms for detecting marine objects using camera, radar, and LiDAR(Light Detection and Ranging), a probabilistic-based data association and tracking algorithm and a new collision risk evaluation method. Because the iSAS estimates motions of all and each objects along with their semantic information, it could not be said as a simple replacement of what the captain does. By sequentially installing the iSAS on a small test ship and a demonstration ship during the project, we will simultaneously perform algorithm development and field verification to achieve reliability in the real environment. The iSAS can be used not only for autonomous ships but also for manned ships to enhance safety and reduce costs in the near future.
Publisher
Institute of Control, Robotics and Systems
Issue Date
2021-08
Language
Korean
Article Type
Article
Citation

Journal of Institute of Control, Robotics and Systems, v.27, no.8, pp.556 - 564

ISSN
1976-5622
DOI
10.5302/J.ICROS.2021.21.0063
URI
http://hdl.handle.net/10203/289516
Appears in Collection
ME-Journal Papers(저널논문)
Files in This Item
122467.pdf(2.81 MB)Download

qr_code

  • mendeley

    citeulike


rss_1.0 rss_2.0 atom_1.0