Recent advancements in robotics algorithms have been made possible by combining probabilistic-based methods with various artificial intelligence technologies. With the support of hardware technologies such as computing power and batteries, autonomous mobility technology has been rapidly advancing across all fields, including land, air, and sea. While marine robots have been somewhat slower to adopt these technologies due to the wide area, various ocean environmental conditions, and their unique characteristics, they have achieved significant advancements in recent years.
This presentation introduces our ongoing research on the situation awareness system of autonomous surface vehicles, including unmanned surface vehicles (USVs) and autonomous ships (ASs). Our work focuses on the development of deep-learning-based object detection and tracking algorithms for cameras, LiDAR, and radar, with data fusion algorithms using mathematical and heuristic methods to identify targets or collision risks around the ship.
During the presentation, we will show some experimental results with two different goal settings for USVs and ASs. We will also discuss the advantages of considering both math-based and heuristic-based approaches when accurately understanding the system and the purpose we are pursuing, rather than strictly adhering to well-established frameworks.
Hyun Choi received his BS, MS, and PhD degrees from the Hanyang University, Korea in 1991, 1993, 2000 respectively. After working for Korea Telecom, Korea and ASL, University of Hawaii, USA as a post-doc, he joined Korea Research Institute of Ships and Ocean Engineering (KRISO), Korea in 2003. He has been leading many projects related to intelligent marine robotic applications including remotely-operated vehicles, autonomous underwater vehicles, unmanned surface vehicles, and autonomous ship. He served a director of ocean system engineering division from 2015 to 2017. His research interests are the design of marine robots, perception, advanced control & navigation using robotic intelligence. Now he is the president-elect of Korea Robotics Society and co-chair of IEEE RAS TC on Marine Robotics.