DARPA demonstrates warfighting force with artificial intelligence as a true partner


The Defense Advanced Research Projects Agency, commonly known as DARPA, has provided details of the program that highlights manned-unmanned teaming to enhance capabilities for ground units, giving squads battalion-level insights and intelligence.

DARPA demonstrates warfighting force with artificial intelligence as a true partner


DARPA’s Squad X Experimentation program aims to demonstrate a warfighting force with artificial intelligence as a true partner. In a recent field test, the program worked with U.S. Marines at the Air Ground Combat Center in Twentynine Palms, California, to track progress on two complementary systems that allow infantry squads to collaborate with AI and autonomous systems to make better decisions in complex, time-critical combat situations.


“We are in a race with potential adversaries to operationalize autonomy, and we have the opportunity to demonstrate autonomy in a way that we don’t believe any nation in the world has demonstrated to date,” said Lt. Col. Phil Root (USA), the Squad X program manager in DARPA’s Tactical Technology Office. “Developing hardware and tactics that allow us to operate seamlessly within a close combat ground environment is extremely challenging, but provides incredible value.”

The exercises in early 2019 in Twentynine Palms followed experiments in 2018 with CACI’s BITS Electronic Attack Module (BEAM) Squad System (BSS) and Lockheed Martin’s Augmented Spectral Situational Awareness and Unaided Localization for Transformative Squads (ASSAULTS) system. The two systems, though discrete, focus on manned-unmanned teaming to enhance capabilities for ground units, giving small squads battalion-level insights and intelligence.

In the most recent experiment, squads testing the Lockheed Martin system wore vests fitted with sensors and a distributed common world model moved through scenarios transiting between natural desert and mock city blocks. Autonomous ground and aerial systems equipped with combinations of live and simulated electronic surveillance tools, ground radar, and camera-based sensing provided reconnaissance of areas ahead of the unit as well as flank security, surveying the perimeter and reporting to squad members’ handheld Android Tactical Assault Kits (ATAKs). Within a few screen taps, squad members accessed options to act on the systems’ findings or adjust the search areas.

Between Lockheed Martin’s two experiments to date, Root says the program-performer team identified a “steady evolution of tactics” made possible with the addition of an autonomous squad member. They also are focused on ensuring the ground, air, and cyber assets are always exploring and making the most of the current situation, exhibiting the same bias toward action required of the people they are supporting in the field.

CACI’s BEAM-based BSS comprises a network of warfighter and unmanned nodes. In the team’s third experiment, the Super Node, a sensor-laden optionally-manned, lightweight tactical all-terrain vehicle known as the powerhouse of the BEAM system, communicated with backpack nodes distributed around the experiment battlespace – mimicking the placement of dismounted squad members – along with an airborne BEAM on a Puma unmanned aerial system (UAS). The BSS provides situational awareness, detects of electronic emissions, and collaborates to geolocate signals of interest. AI synthesizes the information, eliminating the noise before providing the optimized information to the squad members via the handheld ATAK.

“A human would be involved in any lethal action,” Root said. “But we’re establishing superior situational awareness through sufficient input and AI, and then the ability to do something about it at fast time scales.”

The Squad X program has moved quickly through development and is already well along the transition path, due in large part to the program’s focus on partnering with the services to ensure real-world efficacy. For the CACI system, that included an opportunity to test the technology downrange to get real-world information, not simulation. At the most recent experiment with the BSS, service representatives used the system to locate and identify objectives in real time.

For both systems, feedback has included a desire for a user interface so intuitive that training takes an hour or less and any available action is accessible in two screen taps.

Staff Sergeant Andrew Hall with the Marine Corps Tactics and Operations Group (MCTOG), an advisory teammate to DARPA’s Squad X Experimentation program, says the ability to provide early input will guard against developing a product that either isn’t used or is used improperly.

“The feedback process, in conjunction with the actual experimentation, gives the Marines the ability to use the technology and start seeing what it can do and, more specifically, what it can’t do,” Hall said.

With the conclusion of third experiment, the CACI system is moving into Phase 2, which includes an updated system that can remain continuously operational for five or more hours. Lockheed Martin will conduct its next experiment in the fall of 2019.

CACI’s BEAM system is already operational, and the Army has committed to continue its development at the completion of Squad X Phase 2. The Army is set to begin concurrent development of the Lockheed Martin ASSAULTS system in fiscal years 2019 and 2020, and then, independent of DARPA, in fiscal year 2021.