By next June, robotics will have a new cadre.
Pocket-size prototypes called LANdroids, being developed now under the first phase of the Defense Advanced Research Projects Agency’s broad agency solicitation, will act as wireless network nodes that scurry like ants through hostile terrain to keep warfighters in the loop.
Bidders who recently proposed control software or hardware platforms to the Defense Department R&D agency can learn from the experiences of the DARPA robotic Grand Challenge, now renamed the Urban Challenge and in its fourth year. For that timed contest, robot designers struggle to make their driverless vehicles smart enough to navigate autonomously across a hazard-filled landscape.
In contrast, LANdroid designers must make their battery-powered robots just smart enough to get around limited obstacles without losing IEEE 802.11g contact with their nearest fellows. Each must be lightweight, at about 2.2 pounds, and expendable, at a low-volume cost of about $100. DARPA expects the processing power to be on a par with that of ordinary cell phones.
“Nothing like LANdroids is currently in use by the military on the battlefield,” says DARPA’s Jan R. Walker, spokeswoman for artificial intelligence researcher Tom Wagner, the LANdroid program manager.
As dismounted warfighters move into an area, they would drop the little droids along the way. “They would carry many of the mobile nodes,” Walker says, and it’s possible that warfighters will also have to carry a signal-strength meter to set down the droids in advantageous places before moving on.
But the potential for these little nodes beyond the battlefield is obvious: for use by first responders working in disaster areas, by agencies working along remote sections of the U.S. borders and, perhaps most parochially, by any agency team that needs an ad hoc project network where line of sight and signal weakness are an issue.
On their own, the little track-mounted devices would move about on surfaces such as concrete or carpet, while maintaining radio communication across a self-healing wireless mesh network to transmit commands, plans, maps and sensor data. If necessary, they could “tether” to a particular user or sensor nearing the network edge.
In urban locations with competing radio frequencies or buildings that block line of sight, the droids would have to calculate how to improve their transmitting and receiving positions — incidentally saving battery power, which DARPA estimates should last a week or two. Sensors would keep each unit alert to dropping, bumping, direction and movement.
Proposers can start, DARPA says, with the iRobot Create platform from iRobot of Burlington, Mass., plus the Gumstix single-board computer from Gumstix of Portola Valley, Calif., and the Wifistix antennaed expansion board.
Fryske Helms, spokesman for iRobot, declined to comment on the company’s current government and tactical robot programs, which include larger units for surveillance, perimeter patrol, and explosives detection and disposal.
1.1 mph: A self-propelled LANdroid will be capable of traveling at about a half-meter per second — about the speed of a cockroach.
DARPA has specified that all the initial awards for LANdroids must use the existing Mobile Ad Hoc Network (MANET) protocol, designed for mobile wireless routing.
Signal interference from buildings and non-line-of-sight fading and shadowing in complex settings will be unpredictable, so the LANdroids must be smart enough to handle multiple signals in and out.
DARPA intends to award multiple LANdroid contracts. The software awards will run about $1 million per year per effort in three one-year phases, excluding options. DARPA also will award separate evaluation contracts for the hardware.
In the Eye of the Storm
The potential for use beyond battle intrigues the Federal Emergency Management Agency. Although it does not yet have a plan for using droid nodes, says Robert Powers, deputy assistant administrator for disaster operations, “They’re interesting, but we look at all new technologies. We don’t have a specific application yet.”
Wi-Fi, radio, satellites, microwave and even unmanned aerial vehicles are all part of FEMA’s disaster communication picture. Communication is a continual focus, Powers says. FEMA “is working 27 disasters right now, some new, some old. Some don’t impact the communications infrastructure, but you don’t know till the wind starts blowing.”
The agency is aggressively pursuing disaster communications, Powers says, “by building relationships with state and local governments. We can set up emergency wireless service for both data and voice. It depends on the incident — there are a number of standard packages from suitcase size up to tractor-trailer. If we know what’s coming, we can pre-position equipment for the first responders to use.”
But DARPA’s Walker speculates that a LANdroid-like “temporary capability might be useful for civilian first responders during an emergency that destroyed the usual communications infrastructure, or to supplement existing communications for rescue operations inside buildings.”