The U.S. Army’s Robot Tanks Make Great Bait

The U.S. Army is developing remote-controlled robotic vehicles that, the service hopes, will be able to fight alongside conventional tanks, fighting vehicles and infantry as soon as the 2030s.

The service doesn’t plan for any of these Remote Combat Vehicles, or RCVs, to be fully autonomous with their own self-contained artificial intelligence. But it’s not a huge leap to turn a remote-controlled fighting ‘bot—think a ground version of the U.S. Air Force’s Reaper drone—into an autonomous A.I. fighting ‘bot.

After all, autonomous vehicles already exist in the civilian world. Add a gun and teach the A.I. some tactics and, voila, you’ve got a self-steering, thinking tank.

How well would A.I. tanks fight on a chaotic battlefield teeming with human soldiers, manned vehicles and other robots? To find out, California think-tank RAND ran a war game. Among the fascinating findings—remote-controlled tanks make great bait.

That’s not all that RAND’s war-gamers learned. Artillery and reconnaissance—key functions of ground warfare since, well, forever—still matter even after you deploy a bunch of robots on the battlefield.

RAND’s players also determined that remote-controlled vehicles actually are inferior to both manned vehicles and A.I. vehicles in certain key regards. Go with human beings or self-steering ‘bots, but maybe don’t try to compromise between the two.

The RAND war game, the results of which the think-tank published last month, posits a major ground war between U.S. (“Blue”) and Russian (“Red”) forces in the Baltic region in the 2030s. “In each scenario, a Blue mechanized company task force was attacking a Red motor rifle company task force near Gulbene, Latvia, with the objective of neutralizing the Red force and moving through the area to the east,” the players explained.

They ran the scenario twice. The Red force with its T-90 tanks, BMP fighting vehicles, surveillance drones, artillery and infantry was the same in both.

In the first play-through, the RAND experts equipped the Blue company with M-1 tanks, infantry fighting vehicles and infantry plus a few surveillance drones and howitzers and two types of radio-controlled RCVs that the Army is developing for possible fielding after 2023.

The RCV-Light weighs around 10 tons and comes armed with machine guns or rockets in the class of the Javelin anti-tank missile. 

The RCV-Medium weighs around 12 tons and carries a 30-millimeter cannon or similar weapons, essentially matching the armament of a manned M-2 infantry fighting vehicle.

In the second play-through, the RAND experts made the RCVs fully autonomous instead of slaving them in real time to a remote human operator. The experts stressed that the Army doesn’t currently plan to deploy fully autonomous fighting vehicles, although this capability “might be technically feasible in the farther term.”

The main question, of course, was how the remote-controlled vehicles stacked up against their A.I. counterparts and against conventional forces as the Blue company fought through the Red company’s kill zone. The results were unambiguous, the RAND players reported.

Recon and fires still matter

For centuries, armies have deployed scouts to spot enemy troops and fired artillery to drive those troops out into the open. Even in this robotics age, those old methods remain sound. Commanders and planners forget that at their peril.

The relative dearth of surveillance drones and artillery that the RAND gamers gave the Blue company got that company into big trouble, negating whatever advantage it gained from its high-tech robots. “The relatively limited [surveillance] capabilities provided by Blue [drones] enabled concealment for Red, and Red prevailed in several cases by being the first to fire from that concealment.”

“Ubiquitous, advanced [surveillance] and readily-available indirect fires will compel enemy forces to exit concealment and engage.” Those tactics should give the ‘bots a chance to fight.

Remote-controlled vehicles are vulnerable to jamming

“In the baseline game, the need to maintain unobstructed and unjammed line-of-sight communications to the [remotely-operated] RCVs imposed significant constraints on Blue forces, slowing the pace and complicating the management of Blue’s advance,” the experts explained. “In particular, Red’s effective use of backpack jammers (placed before the battle) substantially limited Blue’s ability to use the RCVs.”

Constrained by jamming, the Blue forces sent their RCVs into battle piecemeal. Red forces picked them off one by one. “We identified the following hypothesis for future study: the need to maintain unobstructed and unjammed line of sight to R.O. combat vehicles is a significant limitation subject to exploitation by an enemy.”

Robots make great bait

For decades, one of the main arguments in favor of robotic weapons is that losing them in battle costs money and opportunity, but not necessarily lives. That’s still true today.

“In both games, RCVs were always the first vehicles Blue moved forward and the first to take direct fire,” the RAND players wrote. “In the baseline game, Red decided initially to not engage RCVs, thereby not revealing its forces and enabling them to wait to engage more important targets.”

“However, in the [A.I.] game, Red discarded that tactic given its view that the offensive capability of the autonomous RCVs was clearly greater than that of the R.O. RCVs. Based on player comments, our hypothesis is that the RCVs, whether R.O. or completely autonomous, were used as bait because doing so did not put soldiers at risk.”

“Moreover, loss of the RCVs seemed, at least initially, to not cause Blue substantial regret because the RCVs’ offensive capabilities (especially those of the RCV-Ls) were limited” compared to manned tanks.

In short, if you’re going to field robotic vehicles, make them highly autonomous and don’t hesitate to sacrifice them—that’s what they’re for. But don’t just throw them away by depriving them of recon and artillery support.

The RAND war game could help to shape the U.S. Army’s expanding effort to equip its divisions with robotic vehicles. And it could prove informative for the Russians, too. After all, the Russian army is working on robotic vehicles of its own.