The symptoms of frailty can be detected not only through the clinical measurements of physical or cognitive properties of an older person, but also through behavioral parameters. Patterns of indoor movements of the older person around the house are significant indicators of his/her health status, since they are correlated to the physical condition of an individual (e.g. difficulty of movement due to pain), to clinical issues (e.g. the fact that a person visits the toilet too often), or to the individual’s psychological condition (e.g. depression, leading to very limited movement).

This is why a part of the FrailSafe project involves the development of methods for monitoring the movements of an older person within his/her house. The proposed solution is an indoor localization application.

The indoor localization application is based upon the use of Bluetooth beacons positioned in the older person’s home. Bluetooth beacons are small devices which constantly emit a unique identifier (ID), using the Bluetooth technology. They act as “lighthouses”, sending a signal to any nearby device that is “watching”. The distance from a “watcher” to the beacon can also be estimated from how “dim” the beacon’s signal is at the position of the watcher.

The “watcher” in FrailSafe is a mobile device, such as a smartphone or a smartwatch, which can pick up the Bluetooth signal sent by the beacons. When a mobile device is in the proximity of a beacon, it can read the beacon’s ID and estimate the distance between itself and the beacon. Placing a few beacons in the house and letting the person carry the mobile phone as he/she moves, allows the estimation of the person’s current room, as shown in figure. The older person is supposed to constantly carry the mobile device throughout the desired monitoring period, which normally is several days, in order for any further data analysis to capture significant movement patterns.

A preliminary training phase is performed by the clinical staff before the real data collection with the older person. The clinical person will define the rooms to be monitored and will move randomly in each of them for half a minute. The preliminary phase is necessary as the application needs to learn how the signal strength responds in the different rooms that are monitored.

The indoor localization application is part of the FrailSafe data acquisition methods. The data collected are uploaded to the FrailSafe cloud repositories to be further analysed by other FrailSafe components. Using the localization data, patterns of indoor movements and activities can be computed to extract behavioral features of the older persons. Such features, in combination with different types of measurements from other FrailSafe devices and games, can be valuable for the definition of frailty metrics and the early detection of frailty, which are the ultimate goals of the FrailSafe project.