After more than a decade of development, the Army is poised to integrate autonomous and semi-autonomous tactical wheeled vehicles into its fleets.
An Army requirements oversight council met Feb. 10 to decide if the service will proceed with fielding two kinds of kits and making them programs of record, said Bernard Theisen, program manager for automated ground resupply at the Tank Automotive Research Development and Engineering Center.
Gaining AROC approval will move the technology from the lab to the acquisition community, which will request funding and begin the process of fielding it for noncombat vehicles of all sizes.
It’s an important step in the Army’s vision to expand the use of autonomous vehicles on battlefields, which may one day include tanks and mobile artillery, Theisen said.
“We have demonstrated these things, put tens of thousands of miles on them. We’ve given them to soldiers. We have used them in the desert, the forest, in the rain and the snow. So they have got some good lineage,” Theisen told National Defense.
They have yet to be used in combat, despite more than 10 years of development.
The idea to convert trucks used to haul materiel, fuel and water to self-driving systems emerged at the outset of the Iraq War when insurgents attacked supply convoys with roadside bombs and small arms. As casualties mounted, the Army began looking into leader-follower concepts that would reduce the number of soldiers exposed to ambushes.
At the same time, Congress mandated that one-third of the Army’s vehicles be robotic by the end of the decade, a goal it was not able to reach.
That ideal is still very much in the mind of TARDEC officials, Theisen said. While operations in Iraq and Afghanistan have wound down and there are no longer the numbers of convoys needed to take supplies to forward operating bases, the “threat is still out there,” he said.
The kits, or appliqués, are also seen as a “baby step” for the other concepts. “What we’re really doing is building a foundation for all the future programs,” he said.
That would include ideas such as robotic wingmen, where formations may include a manned command-and-control vehicle that guides unmanned combat vehicles on the battlefield.
“This is going to be the same underlying technology put on other vehicles,” he said. However, there is still a lot of development remaining until military vehicles can operate in those kinds of complex settings, he said.
Meanwhile, military logistics trucks are primed to proceed with the technology, Theisen said. He declined to say ahead of the AROC meeting what technology readiness level the kits had achieved, but generally TARDEC transfers items to the acquisition community when they reach TRL six. The scale goes from one to nine, with nine being fully mature. Commercial driverless technology has now reached a nine, he added.
If the council gives the go-ahead, it will transfer the technology to program executive office combat support and combat service support.
The two kits offer various degrees of autonomy. The by-wire kit offers features that are available today on many high-end luxury cars. It has electronic control of all primary driving functions: steering, braking, throttles and transmission and features adaptive cruise control. That allows it to do passive and active safety. The former alerts drivers of impending danger with lights and sounds. The latter takes control of the vehicle to do tasks such as lane correction, or braking if the vehicle is about to hit something, he said.
“It will slam the brakes on, and do some intervention like kick you back in the lane so you don’t hit other vehicles,” he added.
The autonomy kit does all that plus replaces a driver with higher level cognitive functions: “Where do I go? How do I get there? How to avoid things.” It’s a combination of sensors and high-end cameras and some algorithms that interpret what’s going on around the vehicle, Theisen added.
The autonomy kit allows for remote teleoperation, waypoint navigation and leader-follower convoys.
The lead vehicle could conceivably be without a driver, but currently there is one either in the driver seat or remotely operating it, he noted. All the technologies that comprise the kits are derived from off-the-shelf components, he added.
TARDEC conducted an experiment last year with an eight-vehicle convoy traveling at 45 mph in Georgia. The route was pre-programmed in this case, he noted.
Another experiment involved four tractor-trailers traveling in traffic on Interstate 69 in Michigan. That tested the autonomous appliqué and dedicated short-range communications radios. The Department of Transportation is moving toward a day when these radios are on nearly every vehicle so they can automatically communicate with each other and infrastructure in order to avoid accidents, warn of emergency vehicles approaching or disabled vehicles. Since military vehicles often travel on U.S. highways and interstates, such radios will one day be a requirement, Theisen noted.
One of the factors holding up widespread fielding of the appliqués last decade was cost. The original goal was $30,000 per vehicle. TARDEC has fallen far short of that, and Theisen sees little hope that the Army will be able to move it much beyond $200,000 per truck.
The military, even including the Marine Corps as a partner, cannot take advantage of the economies of scale that the commercial vehicle manufacturers enjoy. Orders have to come in at 100,000 units per year to receive significant price breaks. Until the military can break that threshold, there is little in the way of savings, he said.
The Army has a little more than 300,000 tactical wheeled vehicles and some, like construction equipment, would probably not need autonomy kits. And since the military drives its vehicles for many years, it doesn’t order them in yearly batches like commercial manufacturers. The circuit boards and sensors would have to be replaced every three to five years, but other systems such as electronic controls could last 10 to 20 years, the Army has estimated.
“We just can’t do 100,000 units a year,” Theisen said. “We’re never going to get the automotive radar for a couple hundred bucks.”
He declined to speculate on how many kits would be needed and when they would be purchased.
Sean Bielet, chief executive officer of Endeavor Robotics, said while the car and truck manufacturers may seem ahead of the military when it comes to driverless technology, the armed services have tougher challenges. Civilian robotic vehicles only have to navigate paved roads and streets. Robotic mules and tactical wheeled vehicles that go off road have a different set of challenges, he said.
“You don’t have that in a dismounted troop environment, so obstacle detection and obstacle avoidance become very important,” he said.
“Being safe around people is absolutely critical to the success of these programs,” he added.
The price of sensors needed to guide such platforms overall has come down dramatically over the past decade, he said. For smaller robots, purchases of 2,000 to 4,000 will result in some savings. Software, however, doesn’t garner savings with economies of scale, he noted.
One of the major drivers of lower costs in the robotics industry has been the promulgation of software standards. Companies such as his no longer have to spend a lot of money developing proprietary software, Bielet said.
Theisen said TARDEC adheres to Society of Automotive Engineers standards. The military and industry are looking to develop new standards so robotic systems can talk to robotic systems, he said.
Peter W. Singer, strategist and senior fellow at the New American Foundation, said as far as driverless technology, the U.S. military was the forerunner more than a decade ago when the Defense Advanced Research Projects Agency sponsored a series of three Grand Challenges from 2004 to 2007. Today, it is following the private sector’s lead.
The Army hit the pause button for many different reasons, Singer said. Convoy duties and roadside bomb attacks were major drivers of the technology until the urgency went away.
Singer doesn’t ascribe the high cost of the kits to a lack of purchasing power, but rather to the way the military acquires technology from defense contractors in general.
If the military doesn’t move forward with driverless technology it risks re-experiencing what it did with mobile devices. Warfighters will wonder why they have the technology at home, but as soon as they pass through their base gates, they go back in time a decade.
“As we see more and more robotic cars and autonomous vehicles on the roads, it’s going to push that idea forward” in the military, he said.
One factor may be new leadership at the Pentagon, Singer said.
Secretary of Defense and retired Marine Corps Gen. James Mattis during his career wrote extensively about bringing in technologies to protect troops on the ground. He wondered why they didn’t get the same level of investment as jet fighter pilots. On the ground is where the most casualties are suffered, Mattis has written.
“Equipping ground infantry forces with better kit aligns with his vision,” Singer added.