http://www.eetimes.com/document.asp?doc_id=1325262
When asked where the biggest power savings can come from in a wearable device Behrooz said it’s the sum of the display, processor, communications, sensors and software. In some use cases, such as always-on applications, you have limit the use of communications and GPS which runs down batteries quickly. But if you use sensor fusion to keep most of the battery-hungry circuitry off until needed, you can extend battery life to reasonable lengths.
Saxe added that consumers would prefer to have a battery life of 3-to-6 months with wearables that merely use sensors to report the data a smartphone. But today wearables are just being used by early adopters, for the market to really take off he said we need to have the battery/power issue solved to have wider adoption.
"The top issues depend on whether the developer is thinking of a wearable device without a display, or one with a display. For devices without a display the top challenge is how to get really, really long battery life. The goal seems to be 6 months minimum, with 1 year preferred. For devices with a display the challenge is control and people want non-touch ways to control the device. How to make this ‘intuitive’ i.e. convenient, repeatable and easily discoverable, is an emerging area so people have lots of ideas and need to quickly try them out and refine them," said Saxe
.
To meet the demands of OEMs in terms of power reduction, Stefan said that every part of the application has to be power-reduced to a minimum. Benedetto added that the slowdown in the smartphone market is making OEMs perceive wearables as the 'Holy Grail,' but predicted that it would take at least three years to even get a grip on the power management systems of the future that hardwire functions to "wake up" micro-controllers for the shortest time possible.
The key according to Oh will be smarter algorithms that can batch data for a minimum of on-time for processing.
The key according to Oh will be smarter algorithms that can batch data for a minimum of on-time for processing.
"The next 'big' algorithm for sensor hubs for wearables will be the one that achieves the lowest power," said Oh."Having a DSP doesn’t make you a sensor hub -- you have to have smart software."
Abdi added that batching data is part of the solution, and that their newest devices take that route, but in a hierarchical mode, where some of the data is processed in the cloud and some on the device. He also noted that one of the greatest ways to help is to do background calibration.
The panel also discussed how audio is achieving things, especially when its data is fused with other sensors, to greatly reduce the power needed by wearables. For instance, audio can determine the wearables context more easily, according to Saxe.
However, to achieve the minimum power envelop for wearables, according to Oh, eventually energy harvesting is going to have to enter the mix with all those other ways of minimizing power.
"Energy harvesting will create a whole new generation of products, we’ll see sensors -- even see customized packaged goods. We'll have sensor caps on our teeth to determine how much we eat; determine the calories, the type of food, and harvest energy from chewing," Oh predicted.
Saxe added that the energy harvesting form factor will have to be virtually invisible, sensors in your clothing will become ubiquitous for health issues, such as quickly identifying when a person has a stroke so that medical aid can arrive in minimum of time.MAIL THIS
No comments:
Post a Comment