Northrop Grumman is set to embark on flight tests for its Beacon project this fall, utilizing a modified Scaled Composites Model 437, known as the “Vanguard” jet. This initiative aims to evaluate the performance of artificial intelligence (AI) systems during actual flight operations. The effort underscores a broader commitment from the Pentagon to assess AI technologies in real aviation scenarios before integrating them into future combat fleets.
The Vanguard will operate Northrop’s newly developed Prism software, which is responsible for core flight control and safety functions. This software framework is designed to allow external developers to plug their systems into the aircraft, promoting innovation and collaboration within the industry. To ensure safety during these missions, a safety pilot will be on board to take control if necessary. Initial test flights will focus on validating the Prism software, with further evaluations of partner software expected to commence afterward, continuing into 2026.
In the first phase of the Beacon initiative, six technology companies—Applied Intuition, Autonodyne, Merlin Labs, Red 6, Shield AI, and SoarTech—will participate. Their technologies will be tested across various tasks, including navigation, coordination with other aircraft, and mission planning. This collaborative approach is designed to foster innovation in the field of AI and autonomy, which is becoming increasingly relevant in military aviation.
Launched in June, the Beacon initiative is described as an “open-architecture ecosystem” that facilitates developer connections without requiring them to build their own aircraft. Company officials highlight this framework’s potential to lower barriers for AI firms while ensuring the safety and reliability of the flying platform. According to Dan Salluce, Northrop Grumman’s director of advanced autonomy, there is a significant demand for autonomous testbeds, particularly from companies specializing in AI and autonomy, many of which do not currently produce hardware.
The project aligns with ongoing Pentagon strategies to incorporate AI into current and future aircraft, including initiatives like the U.S. Air Force’s Collaborative Combat Aircraft program, which aims to pair piloted fighters with autonomous drones. The U.S. Navy is also exploring similar ventures in early-stage projects.
Beacon’s Vanguard aims to complement existing Air Force platforms, such as the X-62A Variable In-Flight Simulation Test Aircraft (VISTA) based in California and the Viper Experimentation and Next-gen Ops Models (VENOM) F-16s located in Florida. Unlike these combat aircraft, the Vanguard is designed for more frequent flights at a reduced cost, with Northrop providing its own telemetry and support systems.
This project is seen as an important step towards integrating AI technologies into military operations, with the potential to reshape how future conflicts are managed. Salluce emphasized the mission of the Beacon project, stating that it allows for the integration of AI and autonomy on a platform that can be offered to the latest companies and innovators to demonstrate their capabilities.
The Vanguard itself is a single-seat aircraft with a length and wingspan of 41 feet (approximately 12 meters) and a maximum gross weight of 10,000 pounds (about 4,536 kilograms). It is powered by a Pratt & Whitney PW535 turbofan engine, delivering 5,100 horsepower, with a planned top speed of Mach 0.85 (652 miles or 1,050 kilometers per hour), a range of 2,600 nautical miles (approximately 2,992 miles or 4,815 kilometers), and an operational altitude of 25,000 feet (or 7,620 meters).
