Shield AI software lets Taiwan drone boats shadow intruder

Shield AI has released new details on how its Hivemind autonomy software steered two Taiwanese uncrewed boats through a joint patrol off southern Taiwan, fusing radar and ship-tracking data into a single picture and switching the pair from patrol to escort duty without an operator directing each move. The U.S. company described the July demonstration with Taiwan-based Thunder Tiger Corp. in a case study published this month. Hivemind ran aboard Thunder Tiger’s SeaShark 600 and SeaShark 800, six- and eight-meter uncrewed surface vessels, during a patrol in Dapeng Bay in Pingtung County. Shield AI said the teams went from first software integration to the live demonstration in six weeks, starting with digital models of both boats before testing the software on their actual hardware. “We sought out the world’s strongest technology: Shield AI’s Hivemind. We chose them because we know they have already integrated and tested on 30+ platforms,” said Bryan L., who leads Thunder Tiger’s USV division. At the start of the mission, planners split the patrol area into two sectors and assigned one SeaShark to each. Both boats used maritime radar and the Automatic Identification System, the transponder network that ships use to broadcast their identity and position, to watch the water around them. When a third boat playing a vessel of interest entered the area, Hivemind matched the radar contacts with the AIS reports, which update at different rates and often place the same boat in slightly different spots, and merged them into one continuously updated track shared by both SeaSharks. https://www.youtube.com/watch?vjBARkzFVc1M The SeaShark 600 then fed live electro-optical and infrared video of the target to operators ashore. Hivemind paused the patrol and ordered the two boats into an escort, placing one off the target’s port side and the other off its starboard side to give operators two angles on the vessel and put the boats in position to support an interdiction if ordered. Once the target left the area, the software ended the escort and sent both SeaSharks back to their original sectors. Shield AI also ran a simulation of a larger fleet. Five uncrewed boats using Hivemind’s adaptive patrols covered a 50-nautical-mile (93 km) stretch, compared with 6 nautical miles (11 km) for five boats on fixed patrol patterns, while both groups kept a 95% vessel identification rate. The company said that works out to roughly 8.3 times the coverage with the same number of boats, with real-world results depending on sensors, speed and sea conditions. “We are very optimistic about Hivemind, which brings autonomy to USVs for area search, target tracking, and the execution of these missions,” said William Chen, chairman of Thunder Tiger. The demonstration took place on July 29 and was attended by personnel from Taiwan’s state-run weapons developer, the National Chung-Shan Institute of Science and Technology, and the Coast Guard Administration, Focus Taiwan reported. The two companies signed a partnership in May to bring Hivemind to Thunder Tiger’s uncrewed systems, starting with the SeaShark line, and The Defence Blog covered the public showing of the boats. Shield AI also signed a contract with the National Chung-Shan Institute in February 2026 covering Hivemind tools for Taiwanese-built platforms. Hivemind, which Shield AI says has been integrated on more than 30 platforms including F-16 fighters and drone boats, is entering a market Taiwan is building quickly. The island’s coastline runs more than 1,500 km (930 miles) across the main island and outlying islands, and Taiwan’s Ministry of National Defense has requested funding for 1,320 uncrewed surface vessels along with thousands of coastal reconnaissance and attack drones, according to the Taipei Times. Shield AI said the six-week effort with Thunder Tiger established a repeatable path for installing Hivemind on other uncrewed boats.