Design of a System of Systems (SoS) for the Interoperability of Non-Invasive Sensors for the Care of Older People Carla Taramasco a,b , Fabián Riquelme a,b a Escuela de Ingeniería Civil Informática, Universidad de Valparaíso, Chile, b Centro de Investigación y Desarrollo en Ingeniería en Salud, Universidad de Valparaíso, Chile Abstract We propose the design of a system of systems (SoS) called Seniors Guardian as an integrated environment for healthcare monitoring, in which several non-invasive sensors coexist to monitor falls, nocturia, perseverant behaviors, air quality (carbon monoxide), humidity, and temperature in real time. The modular architecture also allows the integration of new components. This research adheres to the use of Internet of Things to develop ambient assisted living in smart homes. Keywords: Frail elderly, Homes for the Aged, Environment, Internet Introduction A permanent risk in older adults is falls. Around one-third of adults older than 65 suffer a fall each year, and only half can get back on their own after they fall [1]. This can lead to fear, pressure ulcers, osteopenia, loss of muscle mass, dehydration, hypothermia, pneumonia, and even death [2]. The financial costs that countries pay to cover injuries caused by falls in adults over 65 are substantial and sometimes unsustainable over time [3]. Other risky event that may occur inside homes of the elderly is nocturia – the need to wake up at night many times to go to the bathroom. Nocturia is associated with sleep disorders that can lead to reduced concentration, cognitive decline, poor levels of energy, and overall, negative effects on the quality of life [4]. To deal with these issues, a patient-centered care model is recommended, where not only emergency cases are accepted, but also patients are actively committed to their care and therapeutic processes [5]. The advances in information technology (IT) and the reduction in hardware costs favor a preventive and proactive healthcare experience. Non-invasive monitoring of elderly allows for this experience and takes care of early symptoms of neurodegenerative diseases and the generation of treatments, plans, and programs geared towards the needs of this age group. In this context, smart homes quip homes with technology elements (such as sensors, microcontrollers) to monitor actions or events of the residents, as well as environmental conditions at the location, in order to detect emergencies early and reduce the times in making subsequent decisions [6,7]. There are several recent advances in low-cost devices to detect falls [7,8] and nocturia [9] in older people. It is known that people with nocturia are exposed to falls, due to the state of drowsiness at night [10]. However, there are no long-term measures to determine what percentage of falls in older adults are caused by nocturia. To solve this problem, the interoperability of monitoring systems would be necessary. Methods The goal is to develop a platform for continuous monitoring of the adult’s daily activities to detect risk events both inside and outside of home. The key is to provide them with tools while safeguarding their autonomy and independence. The System of Systems (SoS) should contain software and hardware (sensor network) components. The monitoring systems considered are the following: Actimetry (M1): To classify the daily activities of the elderly through infrared sensors of low resolution (for privacy). This will allow generating patterns of daily behavior. After the elderly activity profile is created, a probabilistic model based on Markov chains can determine abnormal behaviors. This module will be used to detect falls with an accuracy of 90%, by means of AI algorithms already in use [8]. Harmful gases detector (M2): To detect abnormal concentrations in environment and thus prevent accidents (such as explosions, inhalation of carbon monoxide) from sources of emission of gases (such as kitchen, stove, heater). Temperature and humidity sensors (M3): To measure environment indicators that could negatively affect the health of the elderly. Alert module (M4): To receive alerts generated by the elderly out of home through a 3G key chain with a panic button. Alerts are sent together with the elderly location to his/her contacts and emergencies. Wandering analysis module (M5): To detect the wandering of the elderly. A mathematical model on the key chain of M4 monitor elderly trajectory and determines if she/he is lost. With the data collected by the sensors, daily, monthly, and annual reports on the activities of the patient will be made. By unifying information from several patients, it will be possible to perform advanced analysis through data mining, to find non-trivial patterns: characteristics of the evolution of old age, disability, dependency, and more in order to guide treatments, plans, programs, and policies. Results We propose Seniors Guardian, an integrated environment for healthcare monitoring (Figure 1). The modularity approach allows the integration of new components. If one module stops MEDINFO 2019: Health and Wellbeing e-Networks for All L. Ohno-Machado and B. Séroussi (Eds.) © 2019 International Medical Informatics Association (IMIA) and IOS Press. This article is published online with Open Access by IOS Press and distributed under the terms of the Creative Commons Attribution Non-Commercial License 4.0 (CC BY-NC 4.0). doi:10.3233/SHTI190540 1572