Entry Date:
July 27, 2011

Wockets: Real-Time Activity Recognition on Mobile Phones


The goal of this open source project is to create software and hardware that permits automatic, 24/7 physical activity and context detection on common mobile phones. We are doing this by iteratively designing and testing Wockets -- miniature, low-cost hardware devices that will measure human motion using accelerometers.

Wockets send data to mobile phones that are processed by software running on the phone to automatically detect type, duration and intensity of physical activity.

The software that uses Wockets is free and open source. The Wockets hardware is designed for low cost, robust operation; all the designs are freely available, and we are doing all we can to make it easy for researchers, hobbyists, and developers to get access to them. For more information, please visit our wiki page: http://wockets.wikispaces.com

The NIH's GEI Exposure Biology Program has granted founding to MIT Changing Places and Stanford Prevention Research Center to jumpstart this effort. Although the medical community is interested in these sensors for long-term health monitoring, our goal is to foster the development of a much wider community of people interested in doing important, interesting, or just fun things with mobile phones that can automatically detect physical activities in real time.