You reach for the peppers, dicing them to the perfect size for your famous homemade stir-fry. As the knife does its work, cameras mounted around your kitchen are watching. They log your knife technique and the size of the dice, while sensors at the stove relay the temperature when you add them to the mix.
All the information is beamed to a computer that crunches the data, essentially learning the recipe—and then storing it online, so it can help your cousin in Florida make sure he’s nailing the steps the next time he busts out the wok for a dinner party.
As crazy as that might sound, the march of cheap, powerful, connected, and increasingly smart gadgets means scenarios like this could be just over the horizon. And a rebooted research partnership at the University of Washington is aiming to decode the path to those experiences.
It’s called the Intel Science and Technology Center for Pervasive Computing, a project that brings $2.5 million per year in direct funding from Intel (NASDAQ: [[ticker:INTC]]). The center is headquartered at the UW, and will be co-led by Dieter Fox, a professor in the UW’s computer science department. Anthony LaMarca, a scientist at Intel, is the other leader. Check out the video below, via Intel, for a glimpse at the vision for the new center.
Fox tells me that the program has at least a three-year commitment, with an option for two more years of funding if Intel wants to continue the investment. The results of the research are intended to be made widely available as published papers and open-source software, Fox says.
Although the UW will be the center’s hub and will have nine people working on the project, researchers from Stanford, UCLA, Cornell, the University of Rochester, and the Georgia Institute of Technology also are part of the core research team. Among the other UW researchers is Shwetak Patel, an expert in sensor networks who just won a MacArthur Foundation “genius grant.”
Fox says the new Intel team is designed to iron out a few wrinkles seen in other types of collaborative research—being small enough to be nimble, but large and diverse enough to tackle wide-ranging subjects like that super-connected smart kitchen.
“Intel put actually quite a bit of trust in the team, where they say, ‘OK, we’ll bring together the best people in the field, and they’re going to do very interesting research.’ And they did not say we should work toward something that could, say five years from now, be a product that Intel could sell,” Fox says.
What Intel does gain (other than some nice press, of course) is insight into the next wave of computing platforms—things that could run on some nice Intel chips.
And that kitchen scenario I sketched out above isn’t just a writing exercise. It’s an example of one practical area that Intel’s research money is targeted toward. Fox says the targets will help the team focus its efforts, which is one key when you bring together researchers from fields as different as low-power sensors and camera-based interfaces.
The new Intel center is a rejiggered strategy for the company’s outside research spending. It previously sponsored standalone labs that collaborated with universities, but switched earlier this year to this new kind of embedded approach, partially to save money on overhead and staffing.
The UW-based Intel center and its counterparts across the country are the keys to a $100 million Intel program to bankroll research at universities. The other centers already announced are working on visual, secure, embedded, and cloud computing research.
