FORT A.P. HILL, Va. — The Army is in the process of testing a new system to allow drivers in truck convoys to keep their hands off steering wheels and their eyes open to spot roadside threats.
The convoy active safety technologies (CAST) system is being developed by the U.S. Army’s Research, Development and Engineering Command’s Tank Automotive Research, Development and Engineering Center based in Warren, Mich.
The program is designed to develop a system at a cost of less than $20,000 per truck.
The goal is to allow trucks or humvee operators who are following a lead driver to stop driving and use their time to scan their environment for roadside bombs or other threats. A secondary goal is to cut down on driver fatigue, said TARDEC engineers working on the project.
“We’re trying to use existing technology to enhance our soldier performance while reducing the threat exposure,” said Edward Schoenherr, project manager of TARDEC’s intelligent ground systems.
The program is a marriage of robotics technology and advanced sensors, fueled by the need to mitigate the damage done to trucks by improvised explosive devices.
Roadside bombs in Iraq and Afghanistan have cost thousands of lives. TARDEC engineers are trying to determine whether taking a driver’s eyes off the road and allowing him to scan his surroundings can lessen the bombs’ impact.
TARDEC brought two trucks to the little-used Fort A.P. Hill — located in northeastern Virginia, about 90 minutes south of the Pentagon — to carry out human factor and technical tests, and display the system to journalists and senior military officials.
The system uses four sensors, computer algorithms, communications links and a box with only a few buttons and lights to carry out its task.
Trucks can be in either “follower” or “leader” modes. Lead trucks do not drive autonomously, however they transmit telemetry data to the follower trucks.
There are only four buttons: On, Off, Leader and Follower. A light that flashes green, amber or red tells the driver how well the system is functioning. Green tells him that there is “high confidence” that the trucks are in contact, amber denotes caution and red means the driver should retake the wheel.
The box on the final system, if it goes into full development, may look different, said engineers with the prime contractor, Perceptek of Littleton, Colo. They may, for example, integrate voice warnings instead of the high-pitched squeal that sounds off when the truck must come to a stop.
In order to keep track of the lead truck, the technology features a global positioning system, two ladars (laser detection and ranging) used for obstacle avoidance, built by the German firm SICK AG, two color cameras and an adaptive cruise control millimeter wave radar.
A communications system shares telemetry data between vehicles.
GPS is used less than 1 percent of the time while the system is on, engineers said. And they would like to reduce that even further. The GPS receivers making it onto the final version is doubtful. Heavy canopy, for example, reduces its accuracy.
“We don’t expect to use it in theater because it is usually being jammed either by us or them,” said David Simon, Perceptek’s program manager.
The goal is to allow the convoy to travel at speeds between 40 and 60 miles per hour on dirt roads.
“Each truck can be a leader or follower, said Bernard Theisen, an engineer with the joint center for unmanned ground vehicles. “So if the front truck gets taken out, the next truck could just take the role of leader.”
The system is not intended to save manpower, he added. There will still be two to three personnel in each truck. “We’re not trying to change the force structure.”
TARDEC brought two Stewart & Stevenson five-ton medium-sized trucks for the demonstrations and test, and collected about 1,000 kilometers worth of data.
Besides the technical capabilities, the “human factor” part of the equation was also extensively tested.
Asisat Animashaun, principal investigator at the Army Research Laboratory in Maryland, said workload, cognition, fatigue, situational awareness and “trust in automation” were examined in detail.
The last aspect is important. Driving down a highway and allowing a robotic system to drive a truck is an unnatural act for many. That may add to driver stress, which could influence the amount of targets identified along the roadway because they’re more worried about the system, she said.
Trust comes in time, she added.
“As the system becomes more robust and they are able to see that we don’t have any failures and everything is going like it should, trust increases,” she added.
Preliminary results looked promising, she noted. Three of four drivers had about a 13 percent increase in the numbers of targets spotted, she said.
Those not using the system spotted about 60 of the 120 targets placed alongside the roads. Some were brightly colored. Others were camouflaged. Those driving hands-free spotted about 75 targets.
However, there didn’t seem to be a remarkable reduction in fatigue. All the drivers reported being equally as tired after the tests, she said. There were no passengers in these experiments.
“All participants were fatigued whether system was on or off,” she added.
The Fort A.P. Hill experiments used only two trucks. For tests slated for next fall at a yet-unnamed location, TARDEC will expand the number and types of vehicles.
One of the main bugs remaining to be worked out is improving the performance in heavy rain. The system has worked surprisingly well in light rain, but has trouble with large raindrops interfering with the cameras, engineers said.
Please email your comments to SMagnuson@ndia.org