Taiwan simulates coordinated maritime drone swarms
Taiwan’s Ocean Affairs Council and National Cheng Kung University have completed a simultaneous flight simulation of two drone fleets totaling 20 aircraft. The system uses multi-agent reinforcement learning to divide tasks, plan routes and coordinate aircraft, while also incorporating leader-follower formations, dynamic obstacle avoidance and a minimum three-dimensional separation of 100 feet, according to 科技新报. Tom’s Hardware reports that the project is modeled around JUMP 20, V-BAT and Flexrotor aircraft and remains in simulation and validation ahead of hardware-in-the-loop testing, prototype integration and maritime flight trials.
The reported simulation demonstrates progress toward AI-coordinated maritime surveillance, but the Ocean Affairs Council says reliability testing is still required before the system can be evaluated for Coast Guard operations.
為了因應中國在台灣周邊海域的灰色地帶襲擾，海洋委員會正與成功大學合作開發無人機群飛系統，目前已完成兩個機群、共 20 架無人機同步飛行模擬，希望透過 AI 自主分配任務、規劃航線及協調飛行，擴大海巡署的海上監偵能力。
根據海委會公布的研究成果，這項計畫由成大航太工程學系闕志哲副教授團隊執行，整合人工智慧、飛行力學及群飛協同控制技術，目前已從單架無人機控制，推進至多個機群之間的協同作業。
#Taiwan Ocean Affairs Council#National Cheng Kung University#Taiwan Coast Guard Administration#Shield AI#JUMP 20#V-BAT#Flexrotor