Date: Sep 27, 2013 Source: (
click here to go to the source)
Honking cars, flashing lights, clicking keyboards. We face a daily bombardment of noises, sights and smells that the thalamus--a relay center for sensory data flowing into the brain--sorts into useful information. That frees up our frontal lobe to ruminate on the important questions ... like which coffee shop to visit.
However, most electronic devices aren't as flexible when it comes to sorting through sensory information. Information has to be parsed by a central processor byte by byte, says Philip Dunham Reed Professor of Electrical and Computer Engineering Mikko Lipasti.
And, processing data constantly is a quick route to a dead battery.
Lipasti and his collaborators think that by drawing some inspiration from our own brains, researchers might be able to build smart devices that can efficiently sort out the environmental information that matters. "The mammalian nervous system is very good at filtering out uninteresting information," says Lipasti. "We're trying to deploy a synthetic version of that into silicon."
Part of a research area known as neuromorphic computing, the core technology is something Lipasti has been working on since 2006. The idea itself would be realized in two parts--a low-power processor for monitoring environmental data and a software interface that other applications can use to listen for specific stimuli. "An application can tell our processor, 'When X happens, wake me up and let me know, because that means that the elderly person has fallen,'" says Lipasti.
He's closer than ever to bringing it to market via Thalchemy, a startup company founded with Atif Hashmi, Andrew Nere and Electrical and Computer Engineering Professor Emeritus Jim Smith.
The National Science Foundation saw potential in the concept and admitted the team to its Innovation Corps program, which provided funding and a seven-week entrepreneurial boot camp where experts asked basic but critical questions about the technology's product potential. "You have to find out if you have anything that anyone would want to buy," says Lipasti.
Insurance and health information companies in particular have shown interest in the potential safety benefits. "For example, we can build a continuous sensor that will know when you get behind the wheel of your car, triggering the phone to disable text messaging," says Lipasti.
The excitement really will start once the tools his team is building are finally in the hands of developers, since they'll be able to use them in ways he can't even imagine yet. "If you can infer the context of a device, you can infer all kinds of cool things," he says.
The team also has received funding from the National Science Foundation Small Business Innovation Research Program, the UW-Madison Graduate School, and the Wisconsin Alumni Research Foundation.
