The aging society of developed countries, the growing number of elderly people in need of supervision results that nowadays several researches focus on developing home activity monitoring systems. The aim of these systems is recognising age-related diseases in an early stadium, as well as detecting emergency situations, e.g. falling of an elderly. With the usage of this kind of system, elderly people can live in their homes for an extended time. The operation of activity observing systems can be based on several types of sensors. Passive infrared (PIR) movement detectors are already widely spread nowadays, they can be found in numerous homes. However, these detectors operate primarily in security or energy management systems, and they are connected to the system with wires. The idea of using PIR sensors for activity monitoring can be raised, but this claims new demands against movement sensors: it has to provide as much information about the activity as possible. Therefore more sensors should be installed in a room, so it is reasonable to implement their communication in a wireless way.
In my thesis I have developed a wireless movement sensor capable of activity monitoring. As the first step, I reviewed which sensor types can be applied to a feasible activity observing, and then I chose appropriate PIR detectors. I analysed in detail, what kind of methods are suitable for extracting activity-related information from the output of the detector, and what kind of demands are claimed by these methods against the hardware. The central element of the device is Atmel’s ZigBit module, it is able to implement Zigbee wireless communication. Beside developing the hardware and software of the movement sensor, I created a simple wireless network system. It contains a ZigBee coordinator device, which collects the data of the sensors, as well as data recording and visualizing softwares running on a PC. Developing the firmware of the coordinator and the software components was also my task. I was able to test the developed sensornode with this system, and experienced, that the measurement results contains relevant information about the nature and intensity of the movements and activities.
The developed sensornode communicates on wireless network and operates on batteries, therefore it is really important to maintain low power consuption. Based on my measurements and calculations, the device can operate for up to 4-6 years on two AA-sized batteries.