DEVELOPMENT OF AN INEXPENSIVE HEALTH MONITORING SYSTEM FOR THE LONE ELDERLY

This paper proposes the design and implementation of a low-cost, low-power and scalable intelligent health monitoring system, e-Guardian, to help the lone elderly live independently. The heart of the system is a base station (BS) that acts as a gateway between cellular network and wireless sensor network (WSN). At seniors' sides are a number of wearable devices (WDs) which are capable of detecting accidental falls, inferring simple activities of daily livings (ADLs), monitoring body temperatures and possibly heart rates, etc. WDs communicate with BS, probably in several hops via range extenders (REs). During an emergency, alert signals will be sent to family members either via SMS or phone call. As data from all WDs aggregates at BS, network congestion will likely occur if every WD sends out massive data. To avoid this, e-Guardian takes a decentralized approach by processing sensor data locally in WDs instead of streaming raw sensor data to BS for processing. An interrupt-driven fall detection algorithm using a digital microelectromechanical system (MEMS) accelerometer has been designed to allow host microcontroller units (MCUs) to process data only upon interrupts and sleep the rest of the time, thus achieving months of standby time. The system consumes very little bandwidth so it can scale easily. The scalability allows e-Guardian to be deployed in not only households but also care centres, hospitals and even entire village/community areas.