San Diego-based Shield AI has successfully completed a second flight demonstration of its advanced mission autonomy software, Hivemind, on a General Atomics Aeronautical Systems, Inc. (GA-ASI) MQ-20 Avenger. This recent test, conducted in a live-virtual-constructive (LVC) environment, marks a significant step forward in the development of combat-ready autonomous systems.
According to a press release from Shield AI, the demonstration showcased Hivemind’s ability to orchestrate coordinated tasks between a physical MQ-20 Avenger and its digital counterpart within a realistic mission scenario. This event builds upon a prior successful integration between Shield AI and GA-ASI during the Orange Flag exercise earlier this year. While the February flight focused on fundamental administrative tasks, the latest June test highlighted more complex tactical capabilities, including formation flying and combat air patrol maneuvers.
Christian Gutierrez, VP of Hivemind Solutions at Shield AI, emphasized the importance of real-world testing for advancing autonomy. He stated that “autonomy doesn’t get better in a lab or inside a PowerPoint. It improves when it flies, when it fails, adapts, and flies again.” This philosophy drives Shield AI’s approach to accelerate the path to scalable, operational autonomy by pushing software into live scenarios.
The demonstration underscored several key aspects of Hivemind’s functionality. Shield AI indicated that the software enabled live-virtual autonomy, where Hivemind commanded both a real MQ-20 and its digital twin, facilitating coordinated actions across both physical and virtual aircraft in a combined mission. Furthermore, the autonomy software seamlessly integrated with command and control (C2) and edge systems for real-time execution. The company also highlighted the open, modular architecture of their system, which features Autonomy Government Reference Architecture (A-GRA)-compliant ground system layers, allowing for platform-agnostic integration and adding realism to the test environment.
Michael Atwood, Vice President of Advanced Programs at GA-ASI, mentioned that this event exemplifies the interoperability and adaptability crucial for future autonomy initiatives. He added that the capability to rapidly integrate and test autonomous elements from diverse vendors ensures that the most effective solutions are available to warfighters, regardless of their origin.
Hivemind, as described by Shield AI, is an AI-powered autonomy software designed to enable unmanned systems to conduct complex missions even in environments where GPS and communications are denied. The software is built for modular integration, operator confidence, and deployment across multiple platforms, providing mission autonomy with proven adaptability and tactical relevance.
