Taiwan funds AI-based maritime drone swarm system — Taipei Times

Taiwan’s Ocean Affairs Council is funding a team at National Cheng Kung University to develop an artificial intelligence-based system that will coordinate different types of drones for wide-area maritime surveillance and long-duration monitoring. This was reported by Taipei Times.

The Council said that creating a multi-tier coast guard surveillance network and enhancing maritime situational awareness are key strategies for ensuring law enforcement at sea and countering various forms of “gray-zone” actions by China.

Coordinating different drones

The research group is led by Chueh Chih-che, an associate professor of aeronautics and astronautics. The team developed an intelligent multi-drone control system by combining AI, fixed-wing vertical takeoff and landing drones, flight dynamics technologies and swarm coordination.

More current news is available on the UA.News Telegram channel Telegram.

The researchers studied the capabilities and control models of three types of drones — Jump 20, V-Bat and Flexroter. To do this, they used three-dimensional modeling, aerodynamic parameter analysis and six-degree-of-freedom flight simulation.

Simulation of twenty drones

According to the Council, multi-agent reinforcement learning technology enabled the drones to independently allocate tasks, plan routes and coordinate their actions in formation. To improve formation flexibility and flight safety, the team also integrated leader-follower control and an intelligent dynamic obstacle avoidance system.

The developers completed simultaneous flight simulations of two groups comprising a total of 20 drones. The control algorithm provides for three-dimensional distance between the aircraft and a minimum vertical interval of 30.5 meters to prevent collisions. The next stage is expected to include hardware-in-the-loop testing, prototype integration and flights over open water in various sea conditions.