The Army will expand its emerging technology experiment this fall, bringing in more operators, more stealth aircraft, Navy standard missiles and new AI tools, and will focus heavily on defeating a high-tech adversary with a striking resemblance to China.
The Army will hold its second Project Convergence experiment at Yuma Proving Grounds in Arizona, and simultaneously at several other locations, from October 12 to November 9, Col. Tobin Magsig, the head of the Army’s Joint Modernization Command, told Defense One. First established last year, Project Convergence has become the Army’s largest technology combat experiment to test out new artificial intelligence, autonomy and software tools even rapid software development under battlefield conditions.
It’s also emerged as the most important U.S. military experiment to test out new concepts for interconnecting planes, drones, ships and operators across the battlefield and across the services, a broad effort called Joint-All Domain Command and Control, or JADC2. Unlike other experiments or military wargames that test current readiness levels, Project Convergence is aimed at rapidly accelerating the Army’s ability to find and take out targets by connecting people, vehicles and weapons through a massive, interconnected sensing and shooting kill web.
Last year’s experiment showed that better data interconnection could significantly speed up the pace of operations, bringing down the time it takes to detect, identify and fire on targets from 15 minutes to less than one. The key to that is quickly sharing lots of intelligence and targeting data across many different weapons, vehicles, operators, etc. But experimental artificial intelligence, such as the Army’s FIRESTORM program, also played a major role in fusing that data and suggesting which weapon or unit was the optimal choice to take out the target.
This year’s experiment will look at those same concepts, but at “distances intentionally built to replicate distances in the INDOPACOM [Area of responsibility.]” said Magsig. Read that to mean across the vast stretches of ocean between Guam, various island chains and China. The experiment will also examine how well different platforms and operators are able to pass information back and forth while a technically advanced adversary attempts to block, intercept and confuse signals.
Magsig didn’t mention China by name, but did say that experiment would further test the military’s ability to breach high-tech anti-access area denial defenses, such as long-range missiles and effective electro-magnetic interference, which China possess to a greater degree than any other potential adversary in the INDOPACOM area. Magsig called anti-access area denial capabilities the military’s greatest challenge in the Pacific.
Project Convergence 2021 will put a bigger emphasis on actual operators using the technology as opposed to scientists.
“It’s … about how does this fundamentally change the way we will fight in the future? And that’s where the operations unit really comes in. I can sit back and imagine it, but to actually put technology in the hands of soldiers in the dirt and ask them to think about how this will enable them to fight in the future, that’s an outcome we will get in 21,” Gen. John Murray, the head of Army Futures Command, said during a Center for New American Security event this week.
The Army’s 82nd Airborne Division, as well as the newly established Multi-domain Task Forces, will play a role. “We’ll have some naval forces off the coast of California in the Pacific,” as well, Magsig said. The Navy will also be participating out of its facility at White Sands, New Mexico, where it will fire off a standard missile 6 as part of the experiment. The Marines will fly stealth F-35 aircraft, and Magsig said he expected the Air Force to do so as well.
The Shadow Air Operations Center at Nellis Air Force Base in Nevada will participate, which Magsig said will help to inform how the Air Force develops air operations centers for future conflict, Magsig said.
In general, he said, this year’s experiment is exponentially larger than last year’s. He described last year’s event as “section level,” with fewer than ten active targets and active units at a time. “This year we will scale it up … to company … [around 200] in terms of the level of targets and also level of autonomous, robotic and semi-autonomous vehicles with humans in the loop that we are asking FIRESTORM to direct and control the fires across a company-sized formation.”
The Army will also be testing how well it can use FIRESTORM “in terms of its ability to interact and in some cases provide decisions for joint fires and controlling joint air space.” He called a “stretch goal.”
Perhaps most importantly, the experimentation will inform the military on how to integrate competing artificial intelligence programs and tools across services and departments under battle conditions.
“The Army is not the only service delving into artificial intelligence and decision-making algorithms. In some cases we may find that the answer is not one algorithm fits all … We may find that the best method forward is to federate and integrate each services’ algorithms so that they interoperate together. So part of what we’re experimenting on in the desert here is it will show us the left and right limits of each of these algorithms and which ones are sort of optimal in different conditions and different situations,” Magsig said.