Personal Robots, Home Sensing, Private Networks, and More from Intel Research Seattle’s Open House
Want wireless power? Better network privacy? Automated elder care? You’ve come to the right place.
That place is the 2008 Intel Research Seattle open house, which I had the opportunity to attend yesterday afternoon. I had gotten a sneak preview the day before from lab director David Wetherall, and just before hitting the demos, I also sat down with Intel’s vice president and director of research, Andrew Chien, for an overview. For the open house, Wetherall noted that Abel Weinrib, Intel’s vice president and director of the corporate technology group, was in attendance, along with representatives from Intel’s business units, and many Seattle-area researchers and industry types.
Then it was time for the fun stuff. Wandering around the sixth floor of Intel’s building near the University of Washington campus, I got a technology-packed tour from the lab’s associate director and principal engineer, Anthony LaMarca. I’ll give just a few of the highlights here. (All photos courtesy of Cheryl Miller at Intel.)
—First, I took in the latest progress in home-monitoring systems for elder care. These include sensors like radio-frequency identification tags on kitchen utensils, which then communicate with a computer to infer when a person is eating, drinking, taking proper medications, and so forth. I covered this area several years ago, and wondered just how far the tech implementation has come. “It’s gone from a vision to something much closer to reality,” said Wetherall. “We’re doing real trials, and sorting out the business value proposition too. We find many parties stand to gain if you do it the right way…Families like it, organizations like it, insurance companies like it. It helps with auditing, as well as providing appropriate care.”
The lab is working with the Veterans Administration on home tests. Home monitoring is related to a broader theme within Intel Research—what it calls “richly communicative” everyday sensing and perception. “Our insight was it was practically impossible to get the deployment right,” says Chien. “And to translate it to a business model was not going to fly.” Chien says the researchers set a “90-90” goal: it should work for 90 percent of activities, for 90 percent of your day. “It’s a coverage goal, and it is really central to achieving large-scale commercialization of the technology,” Chien adds. “This is a second generation of sensing and perception…It’s one of our largest efforts.”
—James Landay, a professor of computer science and engineering at the UW and the previous director of Intel Research Seattle, showed me another example: a monitoring device to help you keep track of your exercise and activity levels, and even what kind of transportation you’re using on a daily basis (walking, biking, driving). An accelerometer and other sensors in the device connect to a processor, which uses your cell phone as a display. Landay says his team is in the process of porting the technology over to the iPhone (which has an onboard accelerometer), and possibly to phones that will run Google’s Android system, because the latter might be a more open platform.
—Intel research scientist Ben Greenstein showed me the “trustworthy wireless” project, which is about improving privacy for users of wireless devices. On a monitor was a map of Seattle showing all the locations his laptop had been broadcasting signals that anyone could use to figure out his identity and where he lives (with software available on the Internet). Another monitor showed exactly what information is sent out when his laptop tries to find a wireless network, or when he opens an e-mail while connected to a network. Greenstein pointed out one nefarious use I hadn’t thought of: a corporate spy might be able to figure out connections between companies and anticipate certain deals just by hanging out in their vicinity. “They might work out if something’s going down,” he says. To defend against this, Greenstein’s software goes in and limits the information being sent out by a device, by working at different levels of the wireless device and network.
—Who knew that Intel works this much on robots? Principal engineer Josh Smith, who did his Ph.D. at MIT with Neil Gershenfeld, showed me a few “personal robotics” projects, including a robotic arm and hand with springy actuators to make it softer, safer, and more adaptive to manipulating objects in its environment (see top photo). Electric-field sensors and a video camera allow it to recognize objects and tell when it is gripping a cup or an apple, say. “Manipulation is the big, hard problem for robotics now,” Smith says. If home helper robots ever take off, I’m thinking Intel wants to be the one to supply their brains.
—Lastly, the most visually striking (and technically speculative) demo was one on “wireless power.” This is the idea that you could potentially charge your phone or laptop without plugging it into a wall socket. Wouldn’t that be something? I didn’t believe it when I first heard about the research at MIT last summer, which was published in the journal Science. (Doing power transfer via electromagnetic radiation has efficiency and safety issues.) But the new idea, which is based on magnetic fields, has held up so far. Alanson Sample, a graduate student in electrical engineering at the UW, showed a demo of a light bulb being turned on by 60 watts of power transferred from one magnetic coil to another, about two feet away (see photo, left). It works by setting up a resonance between the powering coil and the remote coil connected to the light bulb, which gives you an energy efficiency of about 75 percent. Alanson said he’s working on setting up magnetic loops to fit on a laptop. A visitor from laptop-maker Lenovo seemed very interested.
All in all, Intel seems convinced it is getting its money’s worth from its UW research collaborators. “We are the eyes and ears in the community,” says Intel’s LaMarca, who adds that if there’s an interesting idea in the innovation community, the lab makes sure Intel hears about it. On the UW side, the partnership seems to be going well, too. “We’re very excited about the lab being here, and having our faculty members run it,” says Hank Levy, chairman of the department of computer science and engineering at UW. “The lab changes focus every couple of years, but it also keeps some continuity.”