MEMS MotionTracking devices to Enable “Smarter” Wearable Sensing Applications
Over the past several years, the smartphone market has been successful in driving down the size, cost and power of MEMS MotionTracking devices as volume has hit critical mass. Integration has also increased to the point where the MEMS industry has now achieved 9-axis of sensing (3-axis gyroscope, accelerometers and magnetometers) in a single 4x4x0.9mm package. For example, the InvenSense MPU-9150 9-axis MotionTracking device, is simplifying component integration and qualification for wearable sensors and smartphones that require small form factors by providing all necessary motion sensing elements in a single package.
Beyond the motion sensors themselves, a complete MotionTracking solution requires complex sensor calibration and fusion algorithms which, in turn, require high-rate mathematical computation and integration. While smartphones have ample processing power, wearable devices typically run on very low power microcontrollers with limited processing power and memory. To solve this challenge, MotionTracking devices such as the MPU-9150 not only provide sensing, but also embed Digital Motion Processor™ (DMP) hardware capable of algorithm processing that offloads sensor calibration and fusion from the microcontroller while at the same offering embedded intelligence to present interrupt-driven sensor events to the micro that optimize system efficiency and reduce interrupts and power. This embedded intelligence helps system designers optimize wearable sensor power and performance to match the battery life targets for the specific device. The MotionApps™ software platform from InvenSense provides developers the software and algorithms they need to abstract the complexities of managing the sensors and allows focus on higher order system challenges.
So how will MEMS MotionTracking devices make an impact on this market? Some wearable sensing devices on the market today already monitor heart rate or provide pedometer functions. Some even include GPS to track my activities outdoors. Increasingly, the question more people are asking these days in relation to their personal fitness is: “What did I accomplish today?” This is where concepts such as “Quantified Self” and “Contextual Awareness” are starting to gain momentum and where MEMS Motion Tracking intersects fitness and sports via wearable sensing devices.
MEMS MotionTracking devices allow the measurement of linear motion, the rate of rotation, direction and altitude. With this information at our disposal it is now possible with the use of some sophisticated algorithms to accurately classify and recognize any number of activities throughout our day: sitting, walking, running, swimming, jump height, swing speed, or any other physical movement.
With the use of wireless telemetry, this data can be transferred from the wearable sensor to a smartphone or some other hub for processing, analysis and reporting.
David Patterson, known for his pioneering research that led to RAID, clusters and more, is part of a team at UC Berkeley that recently made its RISC-V processor architecture an open source hardware offering. We talk with Patterson and one of his colleagues behind the effort about the opportunities they see, what new kinds of designs they hope to enable and what it means for today’s commercial processor giants such as Intel, ARM and Imagination Technologies.