DOD Funds SBIRs for GPS-Denied Autonomy and AI Training
The U.S. Air Force is funding Small Business Innovation Research (SBIR) contracts for new defense technologies. Recent awards focus on long-range, multi-agent autonomy in GPS-denied environments and AI-driven platforms for creating synthetic data to train both autonomous systems and human operators. The contracts signal the Pentagon's continued investment in robust navigation and next-generation simulation.
- The Small Business Innovation Research (SBIR) program, established in 1982, is a competitive awards-based program that requires federal agencies with significant extramural R&D budgets to allocate a percentage to small businesses to stimulate technological innovation. The Department of Defense (DOD) utilizes this program to address its R&D needs and enhance military capabilities. - AFWERX, the innovation arm of the Department of the Air Force, uses the SBIR program to connect with small businesses and startups to solve national security challenges. AFWERX's Autonomy Prime focuses on accelerating the development and adoption of autonomous technologies, including alternative positioning, navigation, and timing (PNT). - Operating in GPS-denied environments is a major challenge for the military, as jamming and spoofing can disrupt navigation, targeting, and situational awareness, potentially leading to mission failure. Recent conflicts have seen a surge in GPS interference, making resilient navigation systems a top priority. - To counter GPS denial, the DOD is investing in alternative navigation technologies that often incorporate artificial intelligence and machine learning to detect interference and provide accurate positioning by other means, such as analyzing the Earth's magnetic field. - Multi-agent autonomous systems, where teams of unmanned vehicles collaborate, are seen as a way to enhance operational efficiency, extend the reach of manned platforms, and reduce risk to human personnel in complex scenarios. - Synthetic data is crucial for training AI models in defense because collecting real-world military datasets is often expensive, dangerous, and involves classified information. AI-generated synthetic data allows for the creation of diverse and rare scenarios, such as specific threat behaviors or extreme weather conditions, to build more robust and adaptable systems. - The U.S. Army is developing the Synthetic Training Environment (STE), which integrates live, virtual, and simulated domains to create realistic, scalable training for complex warfare scenarios without the logistical costs of large live exercises. - Successful SBIR projects can transition to Phase III contracts, which are funded by non-SBIR government sources to procure the technology for operational use. Companies like Shield AI and Anduril Industries have leveraged AFWERX SBIR funding to secure major follow-on contracts for autonomous systems.
