As populations become increasingly aged, it is more important than ever to promote \Active Ageing" life styles among older people. Age-related frailty can in uence an individual's physiological state making him more vulnerable and prone to dependency or reduced life expectancy. These health issues contribute to an increased demand for medical and social care, thus economic costs. In this context, the INTESA project aims at developing a holistic solution for older adults, able to prolong their functional and cognitive capacity by empowering, stimulating, and unobtrusively monitoring the daily activities according to well-de ned \Active Ageing" life-style protocols.
In this context, the INTESA project is devoted to provide a suite of customizable and highly innovative services to improve the lives and the well-being of elderly and, in general, sedentary or frail people living in their own homes, who su er of illness or several grades of disabilities and are di cult to reach. The suite of services presented in this work is developed using modular programming techniques in order to easily allow further development of tools, algorithms and services. Furthermore, each module is developed using the best state-of-the-art of e-health, e-inclusion, wireless sensors networks and arti cial intelligence applied to data mining. The target users of INTESA will pro-actively and consciously bene t of several services, both in outdoor and indoor environments.
These services deal with most of the user's daily life, in many ways: { promoting physical and cognitive activity both in the domestic environment, using exer-games, and in outdoor scenario, using speci c training schedules. These goals will be reached using wearable devices able to detect physical activity and to provide customized tools to support and verify their tness training; { unobtrusively monitoring the user's sleep quality; { monitoring of changes in weight and level of motor skills and balance. The latter is particularly important in order to prevent falls; { monitoring the protein-calorie consumption and dietary habits; { monitoring social interactions through personal mobile devices and the collection of context information characterising the situation the subject is experiencing (e.g, indoor localization, daily activities organization); { de ning a behavioral pattern model, fusing together short-term monitoring information in order to detect changes in the user's daily routines possibly related to physical and/or cognitive decline, stress conditions and mood.
The rest of this paper is organized as follows: Section 2 describes the existing related initiatives, both in the Italian and European domain, highlighting strengths and weaknesses of the existing approach and how INTESA addresses them; Section 3 presents the core services provided by INTESA and related works in their elds, Section 4 describes the parameters identi ed to assess the well-being of the user and the long-term evaluation plan. Conclusions are drawn in Section 5. 2
The last decade have seen an increasing interest in Health&Well-being research, both from the research communities and from governments. In September 2011 4, the European Commission has published an interesting book, titled \e-Health Projects. Research and Innovation in the eld of ICT for Health and Well-Being: an overview", containing a collection of European projects focused on various 4 http://ec.europa.eu/digital-agenda/en/news/ehealth-projects-research-andinnovation- eld-ict-health-and-wellbeing-overview approaches on this research eld. Furthermore, the European research program H2020 Societal Challenge 1 (Health, demographic change and wellbeing) contains many calls for the study and development of ICT for Health technologies, with the clear objective of promoting an easing technology transfer to the people, especially elderly. Consequently, nowadays it is common to denote this trend using the \Active and Healthy Aging" label and, in this eld, the European commission has set the goal to reach increasing life expectancy and to allow a more independent life for older adults.
Into the European community, the Swedish Institute of Assistive Technology (SIAT), successively Institute for Participation, has o ered signi cant works in the eld, with a focus on remote health care facilities development, especially for people in a fragile state, due to an illness or health problems 5. Furthermore, their fruitful collaboration with industrial partners (i.e. RobotDalen) has provided many technologies, tools, and instruments to the world wide community, allowing prevention of illness, cognitive and mental health issues and, in general, assisted living solutions based upon the conviction that also the people who are facing health issues can lead their life with an overall high quality, both into their own house and in social environments.
From the industry point of view, this ambition has seen an extended participation of the largest high-tech companies. For example, Microsoft and IBM have recently spent a big e ort on this eld o ering many solutions for fragile people. In particular, IBM with the collaboration of Bolzano municipality, has developed the \Abitare Sicuri" project 6, which is able to enable the remote health paradigm. Indeed, residents of this Italian community can bene t from using remote communication with hospitals and clinicians, monitoring their own progresses in the rehabilitation phases. On the other hand, the remote monitoring allows proactive health-care practices and residents can bene t from contextaware solutions, providing early clinical deterioration identi cation and behavioral patterns detection. Finally, these information are used to provide alerts and noti cations to the hospital, expert medical or social services sta , using SMS, e-mail, and online social networks. Partners of this project claims a collection of more than 238000 environmental data and 541 di erent alerts. These alerts have produced a professional sta 's intervention in about the 25% of cases.
Along the guidelines set by the European community, each country has developed its own health-care innovation plan. In Italy, several national project calls are focused on Smart Cities and Communities, aiming at nding people needs and promoting tools for their well-being. In particular, in 7, an interesting work, namely \Italia Longeva", is proposed involving many public organizations (Ministry of Health, Region of Marche, Inrca Institute of Ancona). It manages the health-care challenge from a technical and clinical point-of-view more than the well-being perspective but, at the same time, it represents a good starting point for creating solutions able to connect together several and di erent technologies 5 http://www.mfd.se/other-languages/english/assistive-technology-in-sweden/ 6 https://www-935.ibm.com/industries/it/it/bolzanocity/ 7 http://www.italialongeva.it/ and paradigms, from monitoring of blood glucose to home automation software able to check the contents of the fridge. Another notable project in this eld is represented by \Safe home" 8. It is developed by the Region of Veneto, through \Associazione Temporanea D'impresa" and relies on advanced home automation system development, with a focus on tele-medicine and home-care making possible the integration among hospitals and community services in the provision of care for patients a ected by various diseases. It is based on a medical device that is easy to use, thanks to touch-screen technologies and an accurate user experience interface, and provides information via a privacy-aware cloud infrastructure. Di erent projects focus on speci c technologies enabling the provisioning of remote health monitoring. In this eld, the AMICA project 9, involving several institutes (Istituto Superiore Mario Boella, Universita dell'Insubria, and other private companies), shows an interesting smart-watch, able to detect: i) people falls by employing tri-axial accelerometers and ii) strong sedentariness conditions by analyzing micro ngers movements. The project relies on the use of smartphones and it exploits the Zigbee technology for the data transmission purpose.
Finally, it is worth noticing how these projects and outcomes are centered on a very particular category of people or situation. The lack of a holistic solution that can provide more general services, customizable by users using several physiological, cognitive, psychology aspects, poses a big challenge in this research eld. We deal with this challenge proposing the INTESA system, a modular framework able to provide several services related to the e-health, well-being, and social inclusion. Furthermore, it is highly customizable and expandable, providing services to various category of end-users. This aspect represents its key feature.
The INTESA system is still under development, but it already contains different algorithms and modules able to collect and process sensors data gathered from several di erent sources of information. In particular, INTESA contains modules for sleep quality monitoring, tness and calorie consumption tracker, indoor positioning, posturography and stabilometry monitoring, and it provides tips, suggestions and exercises to the end users in order to improve their own health status and social activity. These modules allow to perform both shortterm and long-term monitoring on which to base the development of techniques for behavioral pattern identi cation, allowing the identi cation of early signs of clinical deterioration in fragile people.
It is worth noticing that each module above mentioned represents a research challenge itself. In the following section we will introduce the speci c issues addressed by each module and the implemented techniques. 8 http://www.consorzioarsenal.it/web/guest/progetti/safehome/il-progetto 9 http://www.adamo-vita.it/
Long-term monitoring services
Nutrition and body mass
Short-term monitoring services Gateways
and User Interfaces
Mobile Social Network
Well-being Primary indicators Users Aggregate data Secondary
Users Well-being Primary indicators Users Well-being indicators extraction
Stress detection & monitoring
Social interaction detection & monitoring Back-end
and storage services Sleep quality assessment
Activity levels
Cognitive and physical assessment
Balance assessment
Indoor position From the architectural point of view, the overall INTESA system is composed of di erent software artifacts that are independent of each other. These modules are deployed on di erent hardware and send the gathered data to a single central unit able to perform the monitoring, detection, and signal processing tasks. In this way, the system is robust to malfunctioning of a single device and it is ready to develop further modules.
Figure 1 shows the overall architecture of the system, together with the distinction between primary and secondary user. The term primary user refers to the older adult who interacts with the system. The primary user is in direct contact with the devices and directly bene ts from the INTESA lifestyle protocol. The secondary user represents the social networks around the primary user, from caregivers and medical sta to relatives, that bene ts from the INTESA indicators to better intervene in helping the primary user to follow the active ageing lifestyle protocol.
The overall INTESA architecture is composed by three major elements: { Short-term monitoring subsystems. These modules provide indicators about the physical and cognitive status of the user, related to the activities performed during the daily living. In this subset we nd: the nutrition and body mass monitoring module; the sleep quality assessment module; the activity detection module; the cognitive and physical exercises monitoring module; the balance assessment module; the indoor position monitoring module. { Long-term monitoring subsystems. These modules provide indicators about the physical, cognitive, and social status of the user, related to the overall period of intervention of the INTESA lifestyle well-being protocol. In this subset we nd: the well-being indicators extraction module; the stress detection and monitoring module; the social interaction and monitoring module by means of a mobile social network. { Back-end services, gateways and User Interfaces. These modules represent the software infrastructure that provides to the di erent subsystems the capabilities to: collect and store data; run the business logic to infer the di erent well-being indicators; provide feedback to primary and secondary users.
In the following sections, we will describe in details the algorithms and technologies used for each module. 3.1
This service aims to monitor and evaluate the temporal evolution of very common pathological conditions in elderly people such as sarcopenia (muscle mass loss/reduction), osteopenia (bone mass loss/reduction), dehydration and excessive visceral fat percentage. The service is based on the use of a bioimpedance scale to collect physiological parameters (e.g., hydration level, body mass index, weight, etc.), and on a mobile application to record dietary habits of the user, based on a personalised diet program.
The service's output represents a valid support and integration for clinical
evaluations by medical sta to control these important pathological conditions
concerning fragile subjects. In fact, dehydration and sarcopenia are considered
risk factors for postural instability, falls, fractures and even cognitive decay
(which can degenerate in Alzheimer disease), and they can often represent
malnutrition indices. Several recent works in literature point out that malnutrition
is one of the most relevant conditions that negatively a ects the health of elderly
people, both in home-based living and assisted living facilities. In almost all the
studies the nutritional status of the subjects is evaluated using the Mini
Nutritional Assessment (MNA) or its short form (MNA-SF) as nutritional screening
tools. One of the most extensive studies presented in [
One of the most important markers of a healthy lifestyle is represented by the
quality and quantity of sleep. These factors directly a ect the waking life,
including productivity, emotional balance, creativity, physical vitality, and the general
personal health. Indeed, poor long-term sleep patterns can lead to a wide range
of health-related problems, such as high-blood pressure, high stress, anxiety,
diabetes, and depression [
Several studies have dealt with the sleep monitoring research challenge [
The sleep quality assessment module of INTESA is able to provide sleep
features in an unobtrusive way, using a small amount of Force Sensor Resistors
(FSRs) and inertial transducers. A similar approach is shown in [
Activity detection and tracking is a research eld that has seen an increasing
interest from the ubiquitous and context-aware computing research community,
which have a considerable impact on the medical settings and protocols [
Considering activity tracking systems, it is possible to distinguish two
different approaches: based on wearable devices and device-free. Nowadays, several
wearable devices have been presented in the market, equipped with several and
di erent technologies and sensors, such as accelerometers as well as biological
parameters transducers. The main drawbacks of this kind of approach is related
to the user experience and comfort. In literature, the need of semi or completely
unobtrusive solutions has been managed using videocameras [
Another key module of the INTESA project is the indoor positioning and
localization. This challenge is one of the main goals of context-aware systems [14{
16], in particular in the Active and Assisted Living (AAL) scenarios [
The hardware of the implemented system consists of: a motion-sensing device, an electroencephalographic (EEG) headset, and a personal computer connected to the integrated database of the INTESA project. As regards the motion-sensing device, we have used the Microsoft \Kinect V2" infrared-based sensor to characterize human gestures and movements. As regards the EEG headset, we have used the Interaxon \Muse" portable system for the acquisition of four EEG traces (Fp1, Fp2, TP9 e TP10) at the sampling frequency of 256 Hz by means of dry electrodes. The electrode mounting is quick and easy because the device is wireless and simply wearable.
The software applications allow the motor and cognitive activities to be executed and the required measures to be performed. These activities are divided into two categories: exercises, aiming to improve the condition of the subjects, and tests, aiming to measure this condition. However, exercises are actually accompanied by measures, and tests can have a positive e ect on the subjects. During the execution of the exercises the movements are constantly monitored by Kinect.
The exercises so far proposed are: mimic the movement of an avatar projected
on the screen; connect the dots; select the tile. The avatar movements present
di erent degrees of complexity; they include movements of the upper part of the
body (e.g. raising arms) and related to lower one (e.g. get up and sit on a chair).
Mimicking the movements is primarily a motor exercise, which however requires
cognitive abilities (with useful e ects reported in literature [
An application for administering a simpli ed version of the classic ANT test
for the measure of the attention level has been implemented. This test consists
of six patterns (a central arrow pointing at right / left with neutral / congruent
/ opposite direction of the ankers); each pattern is visualized on the screen
with a visual angle of 3.8 degrees for 2 seconds with a 4-second interval. The
subject is asked to lift the right / left arm according to the direction of the
central arrow. The e ectiveness of this test [
The objective of this module is the development of a learning system for the
automatic assessment of balance abilities in elderly people. The system is based
on estimating the Berg Balance Scale (BBS) [
In previous work [
The term stressors refers to stimuli of di erent nature that carry the human body and mind to reach high stress levels. Such stimuli can be physical, environmental, social, cultural, psychological, a ective and even nutritional. Stressors can be distinguished in bene ts and harmful. The former ones act as challenges and generate the so-called positive stress or eustress, while the latter ones generate the so-called negative stress or distress, which can lower the immune defenses and lead to rst problems such as anxiety and insomnia. This status can also be followed by long-term disorders and diseases. This service exploits wearable devices focused on the detection of physiological parameters able to re ect the autonomic nervous systems functionalities (both sympathetic and parasympathetic branches). Such parameters include Heart Rate, Hearte Rate Variability, blood pressure, respiration rate, body temperature and Galvanic Skin Response (GSR). The hardware of the implemented system consists of: a wrist unit with nger electrodes for GSR monitoring, a chest strap for cardiorespiratory monitoring and a mobile personal device (smartphone/tablet) which acts as a gateway for collecting sensor data on standard Bluetooth or Bluetooth Low Energy communication channel. As far as the GSR sensor is concerned, we have used Shimmer3 GSR+ Development Kit 10, while the cardiorespiratory monitoring is performed by using Zephyr Bioharness3 11. Both the wearable devices included in the system are equipped with IMU sensors, like 3D accelerometer and gyroscope, to collect physical activity and posture data.
Stressors detection and analysis in elderly people can enriches the monitoring of subjects while executing daily activities, such as physical rehabilitation exercises performed at the gym supervised by physiotherapists, guided cognitive tasks and recreational and social activities. In addition, they represent an important source of information in case of depression. In the current clinical practice, psychiatric diagnosis is carried out through questionnaires and scores scales (e.g. Beck Depression Inventory), ignoring the potential contribution provided by physiological signals. This service is aimed also at investigating how some changes in the nervous system activity can be correlated with clinically measurable mood transitions. 3.7
Maintaining an active social life is very important in elderly to prevent depression and cognitive decline. For this reason, INTESA integrates a social monitoring tool in the services suite. It is based on the use of personal mobile devices, able to exchange data through device-to-device communications (based on BT-LE and/or WiFi Direct standards) and to collect context information characterizing the situation the subject is experiencing. In this way, the application can be used both in controlled environments (like nursing homes) and independently. In the rst case, the application leverages also on contextual information derived from the daily activity organization and by the indoor localization system. In the second case, the environment is more dynamic, and the application is able to monitor also social interactions outside the home environment, both exploiting GPS sensing and by stimulating external subjects to use the same application. In fact, the application is completely unobtrusive, it does not require any interaction with the user and it is optimized for resource-constrained devices. 10 www.shimmersensing.com 11 www.zephyranywhere.com
Well-being assessment and long-term evaluation The services discussed above compose the key modules of the INTESA project. The main goal of INTESA is to build Upon these modules a suite of high-level services that give to the primary and secondary users aggregate indexes of the overall well-being status of the user.
Several works have tried to realize e cient people activity detection, with
a high-level of information [
In the INTESA scenario, we will manage several time-series, gathered from
di erent technologies. Consequently, the INTESA approach fall into the
datadriven category. As shown in [
From the long-term evaluation perspective, INTESA will test its services in a six months trial involving 15 users to be enrolled among the residents of a nursing home co-operating with the project. To the purpose of validation, the study will divide the participants in two sets: the set of users that received the INTESA protocol (intervention group, 10 users) and the set of users that did not receive it (control group, 5 users). The comparison between the improvements achieved by these two sets of users during the pilots was expected to give an indication of the validity of the overall INTESA life-style protocol. 5
The INTESA system is designed to be useful to patient (primary users), clinicians, nurses, and caregivers (secondary users) through several remote monitoring services, improving the coordination among hospitals and community centers. This generates many other bene ts, since INTESA helps reducing avoidable readmissions, o ering more cost-e ective solutions in treating the chronically ill patients, and identifying early signs of clinical deterioration in patients. It can also draws new industrial investment, promoting ICT market opportunities. From a technological point of view, INTESA is composed of di erent modular services covering the main dimensions of the daily living of the users, from monitoring nutrition, indoor activity, and sleep quality to suggesting an \Active Ageing" life-style protocol to the user. The modules are based on beyond the state of the art techniques and infer useful indicators of the physical, social, and cognitive status of the user. From the long-term evaluation perspective, INTESA is planned to be tested in a six months trial involving 15 users divided in two sets, intervention and control groups, in order to validate the proposed approach. Acknowledgments. This work was carried out in the framework of the INTESA project, co-funded by the Tuscany Region (Italy) under the Regional Implementation Programme for Underutilized Areas Fund (PAR FAS 2007-2013) and the Research Facilitation Fund (FAR) of the Ministry of Education, University and Research (MIUR). The authors also wish to thank all the other partners of the INTESA consortium: ESASYSTEM s.r.l. and Kell s.r.l. 44. Palumbo, F., La Rosa, D., Ferro, E., Bacciu, D., Gallicchio, C., Micheli, A., Chessa, S., Vozzi, F., Parodi, O.: Reliability and human factors in ambient assisted living environments. Journal of Reliable Intelligent Environments (2017) 1{19 45. Bacciu, D., Chessa, S., Ferro, E., Fortunati, L., Gallicchio, C., La Rosa, D., Llorente, M., Micheli, A., Palumbo, F., Parodi, O., et al.: Detecting socialization events in ageing people: the experience of the doremi project. In: Intelligent Environments (IE), 2016 12th International Conference on, IEEE (2016) 132{135 46. Palumbo, F., La Rosa, D., Ferro, E.: Stigmergy-based long-term monitoring of indoor users mobility in Ambient Assisted Living environments: the DOREMI project approach. In Bandini, S., Cortellessa, G., Palumbo, F., eds.: AI*AAL2016 Arti cial Intelligence for Ambient Assisted Living 2016. Number 1803 in CEUR Workshop Proceedings, Aachen (2013) 18{32 47. Barsocchi, P., Cimino, M.G., Ferro, E., Lazzeri, A., Palumbo, F., Vaglini, G.: Monitoring elderly behavior via indoor position-based stigmergy. Pervasive and Mobile Computing 23 (2015) 26{42