■ By integrating cross-disciplinary technologies—including autonomous collision avoidance, MIT MOOS-IvP multi-vessel swarm control, Federated Learning image recognition, and GNSS-Denied navigation—the system breaks through the limitations of traditional rule-based navigation.
It enables autonomous learning and real-time decision-making in complex sea conditions while supporting multi-vessel collaborative operations, achieving a fully indigenous, full-domain maritime system for national defense autonomy.
■ JetSea AI holds multiple proprietary patents covering real-time collision avoidance, trajectory guidance, and reinforcement learning-based navigation, along with exclusive licensing of six core technologies from NYCU. The system incorporates MIT's marine autonomy framework with U.S. Navy-validated swarm control capabilities, and has completed paid proof-of-concept projects with domestic partners — establishing high technical barriers that are difficult to replicate.
■ Won the Special Jury Award at the 2025 Satellite Digital Application Innovation Competition. Achieved full autonomous navigation from Kaohsiung to Xiaoliuqiu, including a complete island circumnavigation and return; successfully integrated multiple unmanned vessels to execute collision avoidance and swarm tracking missions.