A prize of $1 million awaits the winning team in Saturday's Grand Challenge, a 200-mile race across the Mojave Desert.
|
|
|||
But the race is no more about the money than Charles Lindbergh's 1927 solo flight across the Atlantic was about claiming the $25,000 Orteig Prize.
This wild scramble by a motley group of robotic vehicles is all about stretching the limits of technology. It's about proving to a skeptical public that machines can "see" and "think" well enough to rapidly traverse a varied terrain.
And it's more than a little bit about pride.
"We don't do this" ---- investing a year's worth of sleepless nights and close to $3 million in cash, borrowed equipment and donated labor ---- "for one race. We are in it because it's who we are," said William "Red" Whittaker, of Carnegie Mellon University, whose Red Team is one of the favorites.
The race sponsor, the Defense Advanced Research Projects Agency, has always been clear about why it is prepared to give away $1 million in taxpayer money. The Defense Department, looking to repeat the success of its unmanned aircraft such as the Predator, is heavily investing in unmanned ground vehicles that will keep human soldiers out of harm's way.
"This is an attempt to accelerate that technology development," said Air Force Col. Jose Negron, who is running the race for DARPA.
According to an analysis last year by the National Research Council, development will have to speed up if the Army is to pursue its Future Combat Systems, a $14.7 billion program that will include unmanned fighting machines, cargo trucks and reconnaissance vehicles.
"It has taken more than 40 years to reach a rudimentary level for unmanned air vehicles, and it is clear that the Army will face at least equally severe challenges to develop [unmanned ground vehicle] systems," the NRC panel concluded.
Major gaps exist in the technology that machines need to perceive their environs, the panel said, as well as in integrating technologies so that machines can act autonomously.
In the next five years, the Army hopes to introduce robotic logistics vehicles, such as cargo trucks, said John Matsumura, associate director for force development and technology at the Rand Corp. But eventually, the Army expects to develop vehicles that will carry weapons.
Gary Schmiedel, director of advanced product engineering for Oshkosh Truck Corp., said the company's military customers had been talking about tele-operation and other automated vehicles for the better part of a decade. That's why Oshkosh has partnered with with Ohio State University as Team TerraMax to enter the Grand Challenge.
"We expect to learn a lot about autonomous control," something that could become essential in obtaining military contracts, he said. The TerraMax is based on Oshkosh's Medium Tactical Vehicle Replacement, a $135,000, six-wheel truck used by the Marines in Iraq and Afghanistan.
Though it is possible to operate some unmanned vehicles by remote control, the amount of bandwidth necessary to control a fleet of vehicles simply will not be available on the battlefield, Rand's Matsumura said. So unmanned military vehicles will need to operate largely autonomously, even though humans would continue to maintain control over firing the weapons they carry, he added.
The Grand Challenge ---- navigating without human guidance for about 200 miles in less than 10 hours ---- is unprecedented, but not unreasonable, based on existing technology and on what will be required for these future military systems, Matsumura said.
"This will help calibrate where we are," he said, giving military officials a better feel for what might be possible.
Dan Kara, president of Robotics Trends Inc., a trade publishing and market analysis firm in Westborough, Mass., said the race might prove more cost-effective than many of the military's previous industrial contracts.
"In terms of bang for the buck, it's great for the military," he said. Consider the intensive effort being made by each of the 24 teams still in the competition: "It's like all-nighter after all-nighter after all-nighter."
And that, he suggested, is bound to produce innovation.
Not everyone is convinced the race will result in any near-term advances, however.
"They're all working their hearts out," acknowledged Joseph Engelberger, often referred to as the Father of Robotics. But the ability of vehicles to navigate across open territory is likely to have limited application, at least outside of the military.
Engelberger, of Newtown, Conn., the founder of Unimation Inc., which produced the first industrial robot arms, regards the Grand Challenge much as he does efforts in Japan and South Korea to develop walking robots.
"It's a nice challenge. But so what? I would much rather see all that talent focused on something that would be useful right now," he said.
Engelberger is a proponent of service robots, such as in-home helpers for the elderly. Navigating across a living room is no big deal for these robots; finding things in a refrigerator, especially items that are hidden behind other items, is a major unsolved problem, he said.
Kara, however, sees the Grand Challenge as "only a good thing for the industry as a whole." Insiders have been keeping a close eye on progress all along, but general news coverage of the race itself is likely to raise the awareness of the public of just how capable mobile robots have become.
"I don't know if you absolutely need a winner," he said. "People are going to realize there's a lot of impetus behind this."