Chinese coast guard researchers have proposed a forward-looking defense concept involving a coordinated network of aerial, surface, and underwater unmanned platforms—dubbed a “robotic fortress”—to be potentially deployed on South China Sea islands and reefs. The goal is to counter the growing threat posed by low-cost drone swarm attacks.
According to Hong Kong’s South China Morning Post, the research was published in July in Command Control & Simulation, a journal overseen by China State Shipbuilding Corporation. The study was jointly conducted by researchers from the China Coast Guard Academy and Dalian Maritime University.
The paper argues that traditional human-centered defense systems struggle against low-cost, saturation-style drone swarm attacks, citing weaknesses in early warning efficiency, cost-effectiveness, and sustained combat capability. It points to Ukraine’s use of Magura V5 unmanned surface vessels against Russia as evidence that such asymmetric tactics can effectively target larger crewed warships.
To address detection blind spots caused by limited island terrain and narrow waterways, the researchers drew on the OODA (observe, orient, decide, act) loop developed by U.S. Air Force Colonel John Boyd, proposing a closed-loop, multi-domain defense system integrating aerial drones, surface vessels, and underwater platforms anchored around islands as fixed nodes. AI-equipped platforms would reportedly retain some autonomous function even if communication links are disrupted.
Separately, satellite imagery reviewed by Reuters shows China has completed the first phase of construction on Antelope Reef in the Paracel Islands, creating a 6-square-kilometer artificial island that analysts believe could become part of China’s largest military base in the region.
The proposal comes amid intensifying U.S.-China military competition in the South China Sea, as both sides seek strategic advantages ahead of any potential future conflict.
Source: Sputnik News, August 20, 2026
https://sputniknews.cn/20260820/1072849620.html