Microsoft and Georgia University researchers have developed a system that can make wearable devices up to nine times faster with four times the battery life by offloading processing to traditional mobile devices.
The platform, dubbed WearDrive, offloads processing power using WiFi and Bluetooth connections so that watches or other wearables crunch only smaller processing tasks.
More ReadingChipzilla has its knockers … and now they’re cool in this venting sports braGoogle sneaks out new clip-on tech goggles for saddo GlassholesMicrosoft open-sources Sora software-defined radioApple's chip'n'firmware security demands behind HomeKit delaysAustralian Research Council funds clean tech
In tests using an Android wearable (akin to the Samsung Galaxy-Gear Smart Watch but with half the RAM) and a phone with a 2000 mAh battery the system improved performance by a factor of 8.85 and battery life by 3.69 with little negative impact compared to state-of-the-art phones.
All told the use of a phone for processing can reduce wearable energy usage by a factor of 15.21x: "WearDrive can leverage phone’s capabilities to reduce energy usage of wearables by up to 15.21x, with trivial impact on phone for realistic wearable workloads", authors say.
It also beat storage systems used in the likes of Google Glass in time lapse photography using its RemoteFlash storage system which stores photos on the paired phone rather than on the wearable using LocalFlash.
Anirudh Badam and Ranveer Chandra. Microsoft.
Video and audio recording is also improved.
Battery life in wearables faces challenges of battery size, the use of power-sucking flash storage, and resource-intensive encryption.
Jian Huang of Georgia Institute of Technology, together with Anirudh Badam, Ranveer Chandra, and Edmund B. Nightingale of Microsoft Research detailed the researcher in the paper WearDrive: Fast and Energy-Efficient Storage for Wearables (PDF).
The team built the WearDrive distributed battery-backed RAM storage system, together with a hybrid Bluetooth and WiFi transfer system, and an application benchmarking app dubbed WearBench to check the performance of devices sharing processing loads.
"Everyone has been thinking of reusing what exists for mobile devices," Microsoft's Chandra says. "What we're saying is, ‘It's a different paradigm. It's a different usage scenario'."
The project is part of a wider Microsoft effort to reduce battery consumption in wearables. ®