Shield AI, ST Engineering join forces on fine-tuning drone swarms

SINGAPORE — American drone company Shield AI plans to integrate its AI-enabled software into Singaporean manned-unmanned-teaming applications, enabling the coordination of drone swarms.

Local firm ST Engineering and Shield AI signed a memorandum of understanding at the Singapore Airshow here on Feb. 5 to combine the Hivemind autonomy software on different platforms manufactured by the national defense-tech champion.

An ST Engineering spokesperson told Defense News the company sees possibilities for integrating Hivemind into its manned-unmanned teaming operating system, or MUMTOS, marketed on the firm’s website as a tool for balancing “human judgment and machine precision.”

The open-architecture system is designed to connect a wide range of unmanned platforms, such as micro-drones, unmanned surface vessels, and robotic ground vehicles.

The Singaporean Armed Forces have been experimenting with collective drone operations. During its largest overseas exercise in Australia recently, the military deployed the largest number of drones in its history, according to the Singapore-based news network CNA. Over 100 drones logged 560 flight hours during Exercise Wallaby.

Hivemind allows unmanned systems to sense and adapt while executing missions by generating new navigation alternatives based on targets and environmental conditions. According to Shield AI, it is designed to continue operating in a GPS- or communications-degraded environment without constant remote control.

The company’s V-BAT drone equipped with Hivemind has been deployed in Ukraine. The AI-intelligence software provides the platform with a global navigation satellite, or GNSS-denied state estimator, that merges sensor data to continue navigation in jammed areas.

“To date, V-BAT has completed about 200 flights in Ukraine [including operational and training missions] – in 2025 alone, it identified more than 200 Russian targets, and we are working with Ukrainian company Iron Belly to integrate Hivemind onto more autonomous strike platforms,” a Shield AI representative said.