No Directions Required--Software Smartens Mobile Robots

DARPA initiative to develop self-navigating robots introduces a world of potential for the development of autonomous vehicles, but will the government take advantage of its research or let it wither on the vine?

By Peter Sergo

SMART ROBOT: DARPA's LAGR initiative awarded each of eight teams of scientists $2 million to $3 million to develop software that would give unmanned vehicles the ability to autonomously learn and navigate irregular off-road terrain.
Courtesy of Yann Lecun, N.Y.U.

Computer experts recently gathered in San Antonio, Tex., to test one last time how well their software programs enabled a mobile robot vehicle to think for—and steer—itself. The event wrapped up the Defense Advanced Research Projects Agency's (DARPA) three-year Learning Applied to Ground Robots (LAGR) initiative, which awarded each of eight teams of scientists $2 million to $3 million to develop software that would give unmanned vehicles the ability to autonomously learn and navigate irregular off-road terrain.

Autonomous maneuvering may not seem terribly difficult for a reasonably smart robot on wheels. But although some vegetation, such as short grass on a prairie, is easily traversable, obstacles such as dense bushes and tree trunks are not. To expediently reach point B, the robot must be able to quickly sort through a range of flora and decide which ones it can travel over—or through—and which are rigid, impenetrable barriers.

Researchers initially believed that visual learning—making basic sense of a surrounding based on changes in light—would be easy to implement in computer systems. But Eero Simoncelli, a principal investigator at the New York University's (N.Y.U.) Laboratory for Computational Vision, pointed out that humans take vision for granted and overlook its complexity. "For you to avoid an object in your path is trivial," he says. "What's visual input [to a computer]? It's a bunch of pixels. It's a bunch of numbers that tell you how much light fell on each part of the sensor. That's a long way from a description of a cup sitting on a table." Extracting symbolic definitions from a large set of numeric values, he adds, is much harder than anyone realized.

Classifying natural obstacles was but one of myriad factors that DARPA researchers had to predict and implement in a software program to expand the capacity of a mobile robot to quickly analyze and travel through an environment. "Of course, no one [knew] how to design this," says Yann Lecun, professor of computer science at N.Y.U.'s Courant Institute of Mathematics who led the university's team. "So DARPA [was] interested in funding projects that advance the science of [robot] learning and vision."

Lecun, who has a knack for designing computer systems that pick out the key visual features in an environment, was an ideal candidate for the LAGR project. DARPA provided the funding and a standard test vehicle so Lecun and Urs Muller, CEO of software maker Net-Scale Technologies in Morganville, N.J., could focus on writing the software. They set out to push the realm of visual-based navigation forward—or to at least bring it up to speed.

A 2002 study by the Washington, D.C.–based National Research Council found that the increase in speed of unmanned ground vehicles was greatly outpaced by the rapid improvement in computer processing from 1990 to 2000 when the physical capability of a vehicle and course complexity is adjusted for. Muller points out that over the past decade there has been a 100-fold increase in computing power and a 1,000-fold gain in memory capacity but developments in unmanned navigational systems have lagged far behind these advances and will continue to without the development of new approaches to visual learning. "The limiting factor in software [design] is the human imagination," he says.

Until LAGR, most self-navigating mobile robots could only scan their immediate surroundings and plot a course over short distances. This made it difficult for robots to figure out an optimum route to any place farther than their own shortsighted universe of about 25 feet (7.6 meters), limiting them to a feel-as-you-go approach that often resulted in time-wasting, circuitous paths to a destination.

Scientists Tuning Very Large Array Radio Telescope for Deeper Exploration

The NSF's Very Large Array radio telescope is getting a digital makeover that will give it the sensitivity to pick up a cell phone signal on Jupiter, and to probe deeper into outer space

By Larry Greenemeier

SILENT VIGIL: The NSF's Very Large Array (VLA) radio telescope has become the Expanded VLA and will be 10 times more powerful when work is completed in 2012.
Courtesy of NRAO/AUI and Laure Wilson Neish

The National Science Foundation (NSF) is in the process of transforming its Very Large Array radio telescope into the—wait for it—Expanded Very Large Array, thanks to digital technology that will boost the Socorro, N.M., facility's already impressive ability to tune in on black holes, supernovae and the rest of the deep space menagerie.

Half of the Very Large Array's (VLA) 28 dish antennas—each weighing 230 tons—have already been upgraded so it can collect eight simultaneous data streams at about two giga- (billion) hertz, up from the previous capability of four data streams at about 50 mega- (million) hertz. The rest of the 28 antennas—which made their debut on the silver screen in the 1997 movie Contact, starring Jodie Foster and based on the eponymous Carl Sagan sci-fi novel—will go digital by 2012, increasing the facility's power 10-fold. The makeover will also replace original components that had been in operation since it was built in the 1970s.

"Certain objects radiate over a wide range of frequency," says Mark McKinnon, project manager for the Expanded VLA. "Improving the sensitivity of the telescope boils down to its bandwidth."

Completed in 1980 but operational before then, the VLA was behind the discoveries of water ice on Mercury; the complex region surrounding Sagittarius A*, the black hole at the core of the Milky Way galaxy; and it helped astronomers identify a distant galaxy already pumping out stars less than a billion years after the big bang.

The increased sensitivity and improved resolution of the EVLA will let scientists peer deep into star-forming clouds and spy on protoplanetary disks of dense gas surrounding young stars as well as track supernovae, fast-moving neutron stars and black holes, McKinnon says. The EVLA's receiving system will be sensitive enough to detect the weak radio transmission from a cell phone at the distance of Jupiter—half a billion miles away—at a projected cost of $94 million.

Data gathered by all 28 of the 82-foot- (25-meter-) diameter dish antennas are brought to a correlator—a central, special-purpose computer—which merges the input into a form that allows scientists to produce detailed, high-quality images of the astronomical objects under investigation. A new fiber-optic system replaces the older waveguide system for taking data collected by the receivers to the central control building and increases the amount of data that can be delivered from the antenna to the new $17-million correlator being built by Canadian scientists and engineers to handle the increased data flow.

In addition to its work for the NSF, the VLA site is also playing an important role in the development of another radio telescope, the Atacama Large Millimeter / submillimeter Array (ALMA). Started in 2003 and scheduled to be completed by 2012 in northern Chile's Atacama Desert at 16,500 feet (5,000 meters) above sea level, the facility employs more than 64 40-foot (12-meter) antennas. Scientists have been using the VLA site to test the performance of the dishes before they are installed at ALMA.

"The observations we make with the EVLA will be complementary with what they do at ALMA and at other radio telescopes," McKinnon adds. "Trying to understand astrophysical phenomena requires a multiwavelength approach."