Wireless system for monitoring and real-time classification of functional activity

The ubiquity of smartphones and their capability to support complex data collection, processing and communications applications is leading to a revolution in personal healthcare. Always `on´ and carried, and seamlessly connected to multiple network technologies such as WiFi, Bluetooth, and cellular carrier, these devices can be adopted as hub for aggregating data sourced from custom body-worn sensors. To be effective, however, these devices must balance the diverse design specifications for intermittent smart phone use with the real-time characteristic of health monitoring telemetry. In this paper we focus on the development of sensing devices supporting in-home and community-based health monitoring. We describe the design of a low powered, compact, wireless, and customizable platform for wearable activity monitoring. Of particular importance in this design is the ability to achieve sufficient spatial and temporal sampling resolution for the activity detection mission while meeting battery size and longevity constraints. The sensor design is paired with data aggregation using a WPAN and smartphone to support a body worn network capable of interaction with local telemetry to provide continuous functional activity monitoring, real time activity classification, and notification. Predicted and measured performance data indicate satisfaction of the mission with huge potential gains after optimization of sampling strategies.