The Defense Department’s Technology Readiness Experimentation is all about finding the technology needles in the haystack.

Through this live-fire and prototype demonstration, T-REX is trying to validate the technical maturation and military utility of specific technologies to fill gaps in the warfighters’ arsenal.

These gaps could be urgent operational needs or from an integrated priority list developed by the goals set by the Joint Staff.

“T-REX is unique in the case that we’re take technologies right out of the laboratory. We’re talking technology readiness levels anywhere between four and six, which is our initial entry criteria. There are special use cases, where we go a little earlier than that, if it’s software use initiatives, but we need to make sure that as prototypes come into fruition from concept and we’re driving the capability and fielding into the warfighters hands, that it works,” said  Lt. Col. Matt Limeberry, the commander of the Rapid Assessment of Prototype Technology Readiness (RAPTR) Task Force in Office of the Undersecretary of Defense for Research and Engineering, on Ask the CIO. “That technology come with their white sheets, their quad charts, and we say, ‘Hey, if you have an unmanned aerial system that can fly an endurance of eight hours and carry a 10-pound payload, well, prove it in any platform, any domain, outside of the laboratory and outside of a controlled environment.’”

Limeberry said his team of about 100 uniformed personnel on the RAPTR Task Force, stationed at Camp Atterbury-Muscatatuck in Indiana, puts these technologies through their paces in an environment that closely duplicates certain operational environments.

“As these technologies come through and use T-REX as a venue, we want to be that publisher, that clearinghouse and underwriting authority for technologies going to the warfighter. That speed of relevancy matters,” he said. “As technologies are being assessed in T-REX, they get that military utility feedback from the warfighter, and can iterate on the spot. That’s purposeful for a reason because it helps us buy down risk now for the future warfighter. What we say is let the warfighter fight tonight with tomorrow’s technology. T-REX does that. It helps us buy down risk. It helps us save costs and iterate technology fast and early, fail often and fast, because when the warfighter needs it.”

Brandon Bean, the senior director for artificial intelligence and machine learning solutions in the chief technology office in the defense division at General Dynamics-IT (GDIT), said his team experienced these concepts in action. He called T-REX  “the ideal proving ground. It created realistic, fast-paced and unforgiving circumstances.”

DOGMA platform put through its paces

At the T-REX exercise in the summer of 2024, GDIT and its partners, including Amazon Web Services, tested out the Defense Operations Grid-Mesh Accelerator (DOGMA) platform, which is an AI software that is aiming to modernize long-range communications to support U.S. air defense systems.

GDIT says DOGMA integrates cloud and AI technologies with satellite connectivity capabilities in order to streamline data processing, analysis and decision making.

Bean said the Army and GDIT used T-REX to validate the platform’s adaptability, latency and ability to perform under pressure, while still being able to show how modular edge AI capabilities can extend decision making to the tactical edge, even without reliable communications.

“We combine resilient mesh networking edge optimized compute and real time AI and machine learning inference to process sensor data without the reliance on the persistent cloud connectivity when necessary, but when we’re in a disrupted, degraded, intermittent and low-bandwidth (DDIL) condition, we’re able to also extend connectivity beyond the edge of the battlefield using AWS as a secure cloud backbone,” Bean said. “It was originally built for NORAD and the Alaskan command to predict aircraft trajectories in low visibility radar denied environments, using historical and real-time sensor data. It was originally built for looking for what we call big hunks of metal in the sky. We quickly were able to adapt the solution for T-REX to adapt to the counter UAS mission.”

Bean said GDIT pivoted its use case toward counter UAS in about two weeks leading up to the exercise and then during T-REX, it modified the technology again in about eight hours to test it at the edge with low-bandwidth.

“In an environment like T-REX,  there’s always something that’s going to be thrown at you. The first was being able to identify when we wanted to design our model retraining triggers. When you’re working on data like UAS data, over time, that data will drift, and that could be from a compendium of things and it’s mainly just the way the data is being sent and we are adding new sensors. So we really were able to get a good sense of what the real world conditions would be, allowing us to essentially model the auto-retraining of our models based on that live data drift and orchestrate how we do that at the edge when we’re actually in detail,” he said. “Another one would be our edge orchestration logic. We’re currently working to harden how DOGMA prioritizes its edge workloads when compute is constrained. That was one of the big things we dealt with out there.”

T-REX is a twice-a-year exercise

Limeberry said during T-REX, the Army has key performance measures to evaluate the technologies against to ensure they meet warfighter needs. These include everything from supply chain risk management to scalability to being able to change, correct and adjust in near-real time.

Bean said DOGMA could be used by other agencies beyond DoD. He said the Department of Homeland Security, for example, is another agency with UAS mission areas, whether it’s Customs and Border Protection or FEMA. The Agriculture and Interior departments also are using drones to measure and manage land and wilderness areas.

“There are a lot of a lot of places out there where people don’t really assume DOGMA is beneficial, but anywhere that you need analytics and you need data moved off of the objective, whether that objective is an internet of things sensor, whether it’s a camera, whether it is edge compute, some type of a radar sensor, DOGMA is applicable,” he said. “It’s that secure backbone that allow you to get that data off of the objective and somewhere that you can use it, whether it is out at the edge or back at the core data center.”

Limeberry said DoD has used T-REX to look at an assortment of other technologies too, including resiliency in that communication, predictability in the AI modeling and deterring, detecting and defeating adversarial aerial platforms that threaten the security of bases.

“We are prefacing into 2026 and we’re looking at what we call ‘Top Gun,’ which is a first-person view UAS platforms and drone-on-drone conflict. We are looking at offensive and defensive swarming capabilities, multi domain, collaborative autonomy, so controlling air, maritime and aerial platforms at the same time through learning autonomous stacks,” he said. “It’s all about how we achieve some of the latest administration executive orders of American drone dominance. These tie into those policies and procedures that we want to achieve in American made and manufactured requirements to build our defense industrial base moving forward into the future.”

Copyright © 2025 Federal News Network. All rights reserved. This website is not intended for users located within the European Economic Area.