As more and more universities and companies set up wireless networks to allow untethered Web surfing, network managers are finding out what a pain it is to maintain all the equipment that keeps such systems running smoothly. Network researchers Steven Wallace and Gregory Travis of the Advanced Network Management Lab at Indiana University have come up with an automated tool that will free wireless-network operators from tedious and expensive hours of manually measuring signal strength and recalibrating transmission equipment. The system marries software to a rotating on-site antenna that periodically measures signal strength in all directions, then remotely makes any necessary adjustments to transmission devices. The Indiana researchers have already built a version of the technology able to manage small-business networks and are working on a more powerful version for university campuses and other large networks. That version will have enough range to help network administrators identify rogue users who are either monopolizing bandwidth or using the system illegally. Wallace and Travis hope to bring this technology to a mobile network near you sometime within the next year.
To make medical devices that analyze tiny amounts of fluid, researchers are building biochips with increasingly complex patterns of channels. At the University of Illinois at Urbana-Champaign, materials scientist Jennifer Lewis and structural engineer Scott White have developed three-dimensional networks of channels that make fluids flow in ways that today’s flat wafers can only dream of. The technology could yield chips for DNA and blood analysis able to handle more complex tasks, circulatory networks that pump chemical glues through self-healing materials, and even tiny but sophisticated chemical reactors. The new biochip’s geometry-channels 10 to 300 micrometers in diameter that look like interconnected square spiral staircases-allows it to mix and process fluids in much less space than current planar chips. To fabricate the chips, a syringe dispenses 16 layers of a special “ink” in staircase patterns across a substrate layer of Teflon. The resulting structure is then coated with resin and heated; the heat melts the ink so it can be vacuumed away, leaving a network of channels that are modified to form mixing towers. The researchers have filed a patent on the technology.
Making Diabetes a Game
A new wireless game could help kids with diabetes stay healthy. Many of them don’t record their blood-glucose levels properly, so they don’t get the right doses of insulin. To play the game, developed at Harvard University and MIT, a kid checks her glucose level with a meter that beams the data to a personal digital assistant, into which she also enters her insulin dosages. The PDA wirelessly transmits that data to a central server and, after the day’s third test, redownloads its most recent records. The child then guesses what her next glucose level will be, based on the trends she notices. Lori Laffel, a pediatrician coordinating a study to test the game at Harvard’s Joslin Diabetes Center, says kids playing the game check their glucose levels more often than those not playing. A larger study is planned for the end of the year, in preparation for commercializing the game.
Shot in the Dark
If a gun goes off in an abandoned junkyard, does anyone hear it? The police do, at least in the handful of U.S. cities equipped with gunshot detection sensors, which listen for weapons’ acoustical signatures and clock the arrival of sound waves to triangulate their origin. But these sensors must be plugged into telephone lines, meaning they can’t be installed in out-of-the-way places, and a separate detector is needed every 300 to 400 meters to produce accurate results. Now New Orleans, LA-based Proxity Digital Networks is testing battery-powered detectors that can be clamped onto trees and poles and that communicate wirelessly with communications towers up to five kilometers away. The Tulsa County, OK, sheriff’s department is testing the system, which transmits information on the location of gunfire to officers on patrol. It can even identify specific types of weapons, which helps police “dispatch a more effective response team specific to the situation,” says Tulsa County sheriff Stanley Glanz.
In more than 30 years in emergency medicine, Larry Miller saw hundreds of patients who needed intravenous fluids or drugs immediately-but didn’t get them because paramedics or doctors couldn’t find a suitable vein. So the CEO of San Antonio, TX-based VidaCare, along with researchers at the University of Texas Health Science Center in San Antonio, developed a device to take advantage of what World War II medics called “the noncollapsible vein”: the bone marrow. After a caregiver injects a local anesthetic, the small, battery-powered drill inserts a hollow needle into the core of the shin bone, just below the knee, with the press of a button. From start to finish, hooking up a patient to fluids takes only 10 seconds or so, Miller says. Surveys show that paramedics can’t start IVs for between 10 and 15 percent of patients who need them, translating to millions who could benefit from the drill each year. Miller expects U.S. Food and Drug Administration approval for the device by the end of the summer.
You’re a tank commander in the U.S. Army. As your battalion approaches a bridge, you see that it’s in enemy hands. Do you retreat, engage the enemy, or try to stop the flow of enemy troops crossing the river? That’s a scenario posed by training software being developed by San Mateo, CA-based Stottler Henke. The software mimics a human tutor, examining the decisions a trainee makes on a simulated battlefield. Unlike existing training programs, the system uses artificial intelligence to interact with the trainee in dialogues where there isn’t one correct course of action. It also adapts to the user’s individual strengths and weaknesses, coming up with questions based on his or her battle plan. Currently, the interface is a keyboard and screen with maps and text, but programmers may design a new speech interface. Initial versions of the software will be ready for use by the army within a year, says Eric Domeshek, the project’s manager. Commercial applications abound, he adds-such as e-tutors for teaching marketing strategy in business schools.
So far, mad-cow disease hasn’t been detected on U.S. soil, but food safety experts are still keen to ensure that the burgers on the grill this summer don’t carry the proteins that cause the disease. A test developed at the University of Arkansas could help. It uses conventional infrared spectroscopy to examine ground meat for contamination with tissue from the central nervous system; such tissue is thought to carry the proteins and is forbidden in meats by the U.S. Department of Agriculture. The process, which could be integrated into meat production lines, is more sensitive than current methods, which sample only small amounts of meat using biological assays to find central-nervous-system proteins. Of particular concern, says Arkansas animal scientist and system coinventor Fred Pohlman, are automated meat recovery systems, which strip muscle from beef bones-including the spine. In a 2002 survey, about 35 percent of meat produced by such systems was contaminated. Pohlman believes that with further development, the infrared scan could turn a potential burger nightmare back into a summer dream.
A computer’s keyboard and screen are merely for interacting with its real guts: processor and memory. So why lug these interfaces around every time you want to take your computer across the hall? You won’t have to, if the “Personal Server” project led by computer scientist Roy Want at Intel Research in Santa Clara, CA, succeeds. Want’s team’s prototype-roughly the size and weight of a deck of cards-has a 200-megahertz processor, a one-gigabyte flash memory card, a rechargeable battery, and a radio transceiver. Put it next to any computer equipped with a wireless card and special software, and the miniserver shows up on the desktop as a separate drive. The project grew out of recent advances in processor efficiency, memory capacity, and wireless transmission speeds, says Want. “We’ve just crossed the threshold. Music, videos, text files-now you can put all of those things in your top pocket.” Intel will commercialize the device once prices for the components hit a “sweet spot,” says Want-which could happen within a year.