Researchers develop system to monitor elderly people in their own homes

Jun 12 2018

Research staff from the Universitat Jaume I in the fields of information technology, psychology and business have developed a monitoring system based on indoor location using mobile devices. The purpose of this system is to improve the observation of the positioning of elderly people, in their own homes and in a non-intrusive way, from Wi-Fi signals.

This new system detects, in the quickest possible way, changes in the behavior of elderly people, which allows the intervention of caregivers or their relatives with all the available information. The procedure does not need any additional infrastructure; it is able to issue warnings before an unusual behavior of the people monitored and allows the caregivers to check this behavior through a web interface.

The study was directed by the lecturer Óscar Belmonte from the GIANT Research Group (Machine Learning for Smart Environments) with the participation of Raúl Montoliu, from the same group; Antonio Caballer, from the Intervention and Evaluation in Socio-educational Contexts Research Group, and Merche Segarra from the Department of Business Administration and Marketing.

According to data from the Spanish Institute of Statistics, referring to the year 2016, 41.7% of people who live alone in Spain are over 65 and 70.7% of them are women. Most of them prefer to live in their own home and not in residences. There are even studies that indicate that, in economic terms, it is better to live in your own home, but this implies that care and attention, in terms of time and economic organization, fall mostly on the family.

"We start from the concept of health and active aging established by the WHO", explains the researcher Antonio Caballer. "In this sense, we talk about the use of new technologies to improve the feeling of loneliness, participation, the level of safety and to keep older people in their homes the maximum possible time, without having to go to senior centers".

Teleassistance has allowed providing care and assistance services to the elderly remotely, in a minimally intrusive way, and in their own homes. This service has been used to detect anomalies that pose an imminent risk to the person's health, such as a fall. With the new system proposed by the UJI team, it will be possible to detect possible risky situations, such as if a person stays in bed more than usual or has not gone to the kitchen to cook.

"Our purpose is to study the evolution in a person's long-term health. If there is a continuous deterioration in their cognitive or physical functions, we should realize as soon as possible and try to determine the moment in which this deterioration began", explains Óscar Belmonte, main researcher of the study.

The system uses objective parameters to determine the usual behavior of the elderly people within their own home, so that it establishes behavior patterns and warns when a deviation occurs. From the signal emitted by Wi-Fi access points, it is possible to build algorithms that, once trained, estimate the user's position according to the intensities measured by a device carried by the user (a mobile phone, laptop, smart watch or any other device with a wireless connection chip). "Once the model of a person's data is constructed, we have to check each day if their behavior coincides with the learned behavior or if there are behavioral changes", says Belmonte.

The behavior pattern is extracted from the data provided by indoor location algorithms. Thus, once the behavior of the person is known, it can be monitored to detect possible deviations from the known pattern. In this way, it will be possible to evaluate if either the deviations are occasional, or they indicate a progressive change in the behavior or a possible risky situation.

The Universitat Jaume I has granted an aid to validate this new system with people over 55 within the StartUJI program for the Valorization of Research Results of the Research Promotion Plan 2017. The project will last 15 months.