The education of Stryker, an 18-ton military monster truck, begins in the warehouse lab of General Dynamics in Westminster, Md.
There, Stryker, one of the U.S. Army's newest infantry vehicles, is fitted with a "ladar" scanner, the equivalent of a mounted pair of eyes that see by emitting 400,000 laser and radar beams and snap 120 camera images every second. Its brain -- a 40-pound computer system tucked inside its body -- processes that data, and makes instant judgments on how to act and where to go.
The eight-wheeled Stryker has already seen service in Iraq as an armored troop carrier with human drivers. The idea is to teach Stryker to accomplish a mission on its own, as a robot. By 2010, robotic Strykers and similar contrivances are slated to be in use as all-purpose battlefield vehicles, surveying battlegrounds, sniffing for land mines, or transporting supplies and troops to the front line.
An unmanned Stryker is part of the military's effort to move more machines into battle to save both money and lives. "Well before the end of the century, there will be no people on the battlefield," said Robert Finkelstein, a professor at the University of Maryland's School of Management and Technology.
Companies throughout the defense industry, among them United Defense LP of Arlington, Lockheed Martin Corp. of Bethesdaand the smaller Gaithersburg-based Robotics Research LLC, are developing robotic systems to fill a variety of military functions. For General Dynamics' robotic systems department, making robot brains -- called autonomous navigation systems -- represents the largest business deal in the unit's 14-year history. In November, it won a $185 million award to develop between 30 and 60 automated-navigation prototypes that can be fitted onto vehicles of different size and function, not just Stryker vehicles.
Creating automated navigation systems for combat vehicles is part of the Future Combat System project to remake warfare. The Army plans to spend $14.78 billion on a new combat system over the next six years, of which autonomous navigation systems is one part, according to Maj. Gary Tallman, public affairs officer for the Army.
Founded in 1990 as F&M Manufacturing, the Westminster plant where Stryker's brain is being developed started out designing small, remote-controlled vehicles. Over time, the 80,000-square-foot facility made robots that sorted mail, read bar codes and packaged pharmaceuticals. General Dynamics purchased F&M, which employs 268 people, for an undisclosed amount of money in 1995.
Using autonomous machines in the military became possible in the mid-1980s, when computer processors became faster. In the 1990s, the development of improved sensor technology allowed machines to pick up more information about their environment. Now, autonomous systems can keep track of their whereabouts using global-positioning satellite links, and talk to comrades and commanders through wireless links that shut off automatically if the signal is in danger of being intercepted.
The first unmanned military vehicles made in the early 1980s by the Defense Department were huge vans the size of UPS delivery trucks, filled with hundreds of pounds of clunky computers that could barely navigate at 5 miles an hour in relatively flat terrain. By comparison, Stryker can navigate through forests and desert environments, or drive on the road at top speeds of 60 miles an hour.
Even with these developments, robots still have a lot to learn.
"Now, we have the basic functioning down, and we're trying to make it smarter at something, or better," said Chip DiBerardino, a senior engineer for General Dynamics who works on programming higher intellect into software.
One recent morning, DiBerardino tested a four-wheeled robot called MDARS (short for Mobile Detection Assessment and Response System), a robotic watchdog that patrols the Westminster lab's snow-covered backyard looking for "intruders." It drives several feet, eyes a parking sign and halts, apparently puzzled, until a human attendant reprograms MDARS to move on.
"Compared to a human, MDARS is really not that smart," DiBerardino says by way of explanation.
Developing a robot is like raising children, researchers say.
Even Stryker's most rudimentary movements require complex calculations that must be "taught" to its brain, using hundreds of thousands of programming codes and mathematical algorithms. When it hits a fork in the road, it selects the gravel route instead of the dirt track. When it finds itself trapped in a cul-de-sac, it backs up to reevaluate alternate paths. In the future, Stryker will learn more tactical behaviors mimicking a human's, like running and hiding in trees or behind hills in the presence of enemies. And if its automated comrades go down, it will learn to request orders to carry out an altered mission.
"We need to work on the nervous system of the robots, so it can really learn on its own by picking up patterns based on its prior experience," said Charles Shoemaker, chief of the Army Research Lab's robotics project office in Aberdeen, Md., which funds robotics research at General Dynamics and at universities and other government agencies.
The Predator unmanned aerial vehicle is the most visible of these efforts to have made it into combat. It debuted in Afghanistan and Iraq, collecting aerial images and sending them back to the home base. But autonomous navigation -- allowing an actual unmanned land vehicle that thinks for itself to rove into battle situations -- is a taller order. It requires maneuvering around obstacles, ditches, signs and traffic, which are harder tasks to teach a machine.
Now that it can see and move, Stryker needs to learn how to perceive more and plan better, said James Albus, a senior fellow and researcher at the National Institute of Standards and Technology in Gaithersburg, which has helped develop some of the intelligence used in Stryker's brain.
"In a way, we're trying to duplicate the processes in the brain, and the brain's got a lot of little computers."
One recent morning, DiBerardino tested a four-wheeled robot called MDARS (short for Mobile Detection Assessment and Response System), a robotic watchdog that patrols the Westminster lab's snow-covered backyard looking for "intruders." It drives several feet, eyes a parking sign and halts, apparently puzzled, until a human attendant reprograms MDARS to move on.
"Compared to a human, MDARS is really not that smart," DiBerardino says by way of explanation.
Developing a robot is like raising children, researchers say.
Even Stryker's most rudimentary movements require complex calculations that must be "taught" to its brain, using hundreds of thousands of programming codes and mathematical algorithms. When it hits a fork in the road, it selects the gravel route instead of the dirt track. When it finds itself trapped in a cul-de-sac, it backs up to reevaluate alternate paths. In the future, Stryker will learn more tactical behaviors mimicking a human's, like running and hiding in trees or behind hills in the presence of enemies. And if its automated comrades go down, it will learn to request orders to carry out an altered mission.
"We need to work on the nervous system of the robots, so it can really learn on its own by picking up patterns based on its prior experience," said Charles Shoemaker, chief of the Army Research Lab's robotics project office in Aberdeen, Md., which funds robotics research at General Dynamics and at universities and other government agencies.
The Predator unmanned aerial vehicle is the most visible of these efforts to have made it into combat. It debuted in Afghanistan and Iraq, collecting aerial images and sending them back to the home base. But autonomous navigation -- allowing an actual unmanned land vehicle that thinks for itself to rove into battle situations -- is a taller order. It requires maneuvering around obstacles, ditches, signs and traffic, which are harder tasks to teach a machine.
Now that it can see and move, Stryker needs to learn how to perceive more and plan better, said James Albus, a senior fellow and researcher at the National Institute of Standards and Technology in Gaithersburg, which has helped develop some of the intelligence used in Stryker's brain.
"In a way, we're trying to duplicate the processes in the brain, and the brain's got a lot of little computers."