Original URL: http://www.theregister.co.uk/2003/11/16/darpa_chisels_little_guy_out/

DARPA chisels little guy out of $1 million race

Tethered to Pentagon pork

By Ashlee Vance

Posted in Business, 16th November 2003 17:37 GMT

A few hundred people scattered around the U.S. have embarked on a strange journey that includes robotic vehicles, the government's wackiest research body, a $1 million prize and now a turn of events seen by some as scandal and by others as common sense.

DARPA (Defense Advanced Research Projects Agency) kicked off the Grand Challenge event in 2002 with the best of intentions, but of late questions have popped up as to whether the contest is living up to its original ideals. The plan was to let the Average Joe have a crack at out-inventing military contractors such as Lockheed Martin and Northrup Grumman, while saving taxpayers millions. Teams made up of self-taught engineers, university researchers and small companies were to design robotic vehicles that could see, steer and maneuver on their own across the Mojave Desert in a race like no other.

Excitement around the contest bubbled right up into religious fervor. DARPA dangled a $1 million prize in front of the contestants, promising the cash to whomever could complete a 200 mile race first in under ten hours. The Defense Department arm hoped this motivation would create some fresh insights in robot technology, as its lucrative contracts handed out to the military complex had generated few successes. The idea was to tap into new ways of thinking and let the little guy have a chance at becoming a technology leader.

As a mid-October deadline for contestants' technical papers drew near, the buzz around the Grand Challenge hit an all time high. Over the course of two weeks, DARPA was flooded with last minute entries to the contest. In fact, the quantity of entries overwhelmed DARPA, throwing the organization into a spell of utter confusion.

As discussed here, DARPA began changing the contest rules at will. It had originally proposed that any team could participate in the March of 2004 race, if their technical paper passed muster. DARPA, however, was not prepared for the number of papers, and this approach became unfeasible for a couple of reasons.

For one, DARPA is concerned that branches of the U.S. government and green loving hippies might have a bad reaction to myriad vehicles plodding across the desert. The vehicles come in all shapes and sizes from ATVs to Hummers and, well, don't have any drivers. Each robot is led by a series of GPS coordinates and is equipped with an emergency stop device, but that's where the control ends.

Secondly, the best accounts from race participants suggest DARPA does not have the resources to judge each paper on a quick enough clock.

So instead of looking over all of the papers and then making its final picks, DARPA decided to approve 19 teams for the race who had submitted solid papers early on in the application process. Guaranteed spots.

It then threw out a chunk of papers, saying the teams had no chance to compete and labeled a group of 26 teams as "possibly acceptable." These 26 teams can now request an on-site visit from DARPA and fight for one of six remaining spots for the actual race.

In a situation like this, my dad - a patient man - would say, "Ain't that a heck of a note."

Many of the contestants who have coughed up their proverbial blood, sweat and tears for the last few months see the decision in a much different, more impatient light.

A lot of the small teams were depending on being approved for the race to secure additional funding for their projects. If their technical papers passed the test, the teams hoped to attract sponsors with the promise that they would compete in the race. They were, in some ways, a slave to the way DARPA had set the contest up, and then the agency went and pulled a fast one of them.

On a very serious note, the teams are also concerned that DARPA granted itself a peek at their intellectual property via the technical paper without offering a fair shot at competing in the race. DARPA has first access rights to all of the technology.

DARPA has nudged the contest in favor of the well-funded, experienced teams with its last minute rule change, and a lot of contestants are blaming agency Director Tony Tether for this shift. Tether was tapped in 2001 to take the helm of DARPA by friend of the little guy Defense Secretary Donald Rumsfeld. Scared yet?

It turns out that Tether has been dealing with the military complex for years. He started a company called The Sequoia Group back in 1996 that "provided program management and strategy development services to the government and industry." He has also held several positions within the Pentagon, been on the Army Science board and worked at a variety of other companies. He is also the chap that gave the go ahead for the Total Information Awareness project and an Arab hunting roulette wheel - both projects were nixed after sufficient public outcry.

"This change is coming directly from the top, the Director of DARPA, Dr. Anthony Tether," wrote one Grand Challenge racer who asked to remain anonymous. "Dr. Tether is indeed turning the Grand Challenge into a political event."

A number of letters came in with a similar sentiment. DARPA declined repeated requests for an interview, sending along a press release about the event instead.

The recent rule changes by DARPA do send a disconcerting message.

During an early conference on the event, Tether made the following remarks.

"We attracted a diverse mix of disciplines and personalities today, and that reflects the spirit of the Grand Challenge program," said Dr. Tether. "By bringing together leaders in business, defense, technology and academia with nontraditional partners in fields such as robotics, entertainment and off-road racing, we will develop synergies that foster new ways of thinking."

But many feel the "new ways of thinking" part of the equation has been rubbed out.

Approved teams such as CalTech and Carnegie Mellon University are actually receiving significant help from military contractors that have scooped up DARPA funding in the past. In the case of CalTech, Northrup Grumman is providing some funds and a generous amount of consulting and technology.

Carnegie Mellon's situation is more complex. In the past, the school has competed against Lockheed Martin on robotics projects. Now, however, Carnegie Mellon has formed a partnership with Rod Millen Special Vehicles - a company that specializes in custom engineering projects. Rod Millen and Lockheed Martin have extremely close ties. The two companies tapped into a grant handed down by DARPA in 2002 to build, you guessed it, a robotic vehicle.

Some of the smaller teams are crying foul at these connections. They charge that DARPA is now indirectly funding Carnegie Mellon for the project. This is a big no-no in a competition meant to be open to the public and void of government assistance.

To be fair, a vast amount of work being done at CalTech and Carnegie Mellon is coming from the efforts of undergraduate students more interested in beer than the military complex. In addition, both of the project leaders at the respective schools are working as hard as any other team leader out there.

Still, the links back to DARPA contracts are disheartening for the small teams, particularly with rule changes that benefit well-funded teams. Those with cash and equipment already on hand had a much better shot at completing a top notch technical paper. But the technical paper was originally meant to be just the start of the process - not the end point.

Even teams such as the Arctic Tortoise in Alaska have questioned DARPA's actions. Arctic Tortoise was one of the original 19 teams to be approved for competition but considered boycotting the race given DARPA's "flippant disregard for the competitors’ significant investment."

There are plenty of teams that argue the smaller guys just were not prepared for the race anyway. If you are still waiting for funding at this point and don't have a vehicle ready to go, you have no chance of winning the race.

But this logic seems to fly in the face of the race's original tenets.

Think back to a pair of inventors known as the Wright brothers. With but a few thousand dollars they came up with innovative testing techniques, propeller designs and a light plane that their well-funded competitor Samuel Langley could not match. Langley, you might recall, received his tens of thousands of dollars from the U.S. government and his Great Aerodrome crashed into the Potomac River "like a handful of mortar." By contrast, the Wright brothers came to symbolize a willingness to face the impossible with relentless entrepreneurship that is, melodrama aside, fastened to the American spirit.

It might be best to stick with the little guy as promised in this case. The competitors have proposed various ways of letting more teams compete, while still adhering to environmental and logistical concerns. Many of the teams are now also talking about holding their own, private race possibly in Mexico.

DARPA has already changed the rules a few times, maybe it's time to change them once again. ®

Related Link

Grand Challenge Web site

Related Story

DARPA's indecision threatens integrity of $1 million race