March 3, 2015 -
Electronic parts firm ROHM has partnered with Kobe University Graduate School of System Informatics to develop the world’s smallest ultra-low-power technology optimized for wearable biosensors, according to a report by My News Desk.
Professor Yoshimoto Masahiko of Kobe University’s department of information science is working with ROHM for the ‘Development of Infrastructure for Normally-Off Computing Technology’ project, which is being conducted by the New Energy and Industrial Technology Development Organization.
The technology uses non-volatile memory in order to suppress standby power generation by enabling the power supply to be turned off when no processing/operation is required.
Typically, the off operation is achieved by turning the power on immediately when it is required, which reduces overall power consumption.
By adopting this technology, wearable biosensors are able to easily measure the heart rate by spotting small potential discrepancies on the skin’s surface (EKC/ECG waveforms) using only 6uA of current.
Other digital sensors such as accelerometers can collect, measure and record additional data.
Communications functions are built in that enable biosensor control capabilities via a mobile device and to facilitate data input/output. All operations only require 38uA, which is five times lower than conventional products
NEDO launched the ‘Development Infrastructure for Normally-Off Computing Technology’ project in 2011 to achieve greater energy savings by using non-volatile devices that collect data even when the power is switched off.
The department also conducts R&D for other projects with support from various companies and universities. Among these, ROHM and Kobe University’s Yoshimoto Laboratory have benefited significantly from non-volatile logic technology and ROHM’s non-volatile memory to establish proprietary normally-off technology.
They are currently adopting this technology for wearable biosensors using ROHM’s 0.13um CMOS processes.
ROHM implemented three power-saving measures to achieve the lowest power consumption in the industry, including reducing the power consumption of heart rate acquisition block by 20-fold, decreasing the average power consumption of the memory block to less than 1/10th conventional levels, and lowering the power consumption of the logic block by more than half.