AmIE

Towards Ambient Intelligence for the Ageing Citizens

Julia Kantorovitch

, Jouni Kaartinen

, Luis Collantes Abril

, Ricardo de las Heras Martín

José Alberto Martínez Cantera

, Johan Criel

and Marcel Gielen

VTT, Vuorimiehentie 3, 02044 Espoo, Finland

Indra, 28223 Pozuelo de Alarcón (Madrid), Spain

Telefonica I+D, 47151 Valladolid, Spain

ROBOTIKER-Tecnalia,Parque Tecnológico de Bizkaia, edif. 202, E-48170, Zamudio, Spain

Alcatel-Lucent, Copernicuslaan 50 2018 Antwerpen, Belgium

Mextal the technology company 5674TB Nuenen, The Netherlands

Keywords: Ambient Intelligence, elderly, home environment, assistance, adaptive interaction.

Abstract: This research presents a system, currently under development, which aims at providing an intelligent

ambient able to improve the quality of life and delivering customized support to elderly people in need of

assistance, according to their own specific situation, and in a non-intrusive and respectful way.

1 INTRODUCTION

Celebrating a 100

birthday is a privilege only

reserved for a happy few. However through modern

technology and medical progress, these ‘few’ have

grown to a bigger group than ever before. People

older than 64 in the euro area are about 18% of the

total population in 2005 and it’s supposed to grow

up to 30% by 2050 (Eurostat Yearbook, 2008). Thus

this situation has favoured the need to exploit

technologies that address the issues posed by

demographic ageing and trying to decrease the costs

of health-care systems because so far it seems very

high. Technologies that can support independent

living at home rather than being institutionalized are

envisaged as one of the main solutions for a

significant reduction of health-care costs.

The research described in this paper has

addressed these needs developing a comprehensive

reliable adaptive distributed service system that can

serve multiple categories of elderly people, assisting

them in their everyday life, helping them to stay

independently at home as long as possible and

moreover, increasing their comfort and safety and

social wellbeing. In following, the related research

addressing aging citizen’s needs (with position of

our research in the field) are discussed in chapter 2.

The intelligent distributed service system and

targeted applications are presented in Chapter 3.

Market opportunities are discussed in Chapter 4.

Finally, Chapter 5 concludes the paper identifying

further R&D challenges still faced by researchers.

2 RELATED RESEARCH

The use of information technologies for assisting

people with medical needs and for enabling ageing

citizens to manage their well-being and safety has

been considered in the recent past. Various existing

initiatives can be divided in four big groups:

• Robots

• PERS-systems

• Domotics and sensor networks

• Video telephony and iTV

With the rise of more advanced robots, their

usage for elderly care has increased slowly. In many

of the projects however, the robots, which were

created for a more generally purpose, are being

reused for the elderly (Shinozaki et al., 2005), which

may cause the problem. The main advantages are

that a robot can give the elderly an immediate

feedback on his or her actions, reacting on the spot

of external stimuli. Furthermore, a robot can also be

used to execute physical work of which the elderly is

421

Kantorovitch J., Kaartinen J., Abril L., Martín R., Cantera J., Criel J. and Gielen M. (2009).

AmIE - Towards Ambient Intelligence for the Ageing Citizens.

In Proceedings of the International Conference on Health Informatics, pages 421-424

DOI: 10.5220/0001433404210424

 SciTePress

not capable anymore. On the other hand, robots can

not replace human touch in everyday life assistance.

PERS (Personal Emergency Response Services)

are the best-known elderly assistance systems at the

moment (Porteus and Brownsell, 2000, Kenchiku,

2005). The system acts as an emergency button for

elderly to press when they are in trouble. Also more

automated solutions with activity monitoring

features are available. The great advantage of the

PERS systems is that they are a relatively easy to

deploy as use.

Domotics and sensor networks have started its

uptake about a decade ago (Fellbaum and Hampicke,

2002, Van Berlo and Fellbaum, 1999., Van Berlo,

2005, MOBILALARM, 2005), but because of its

high purchase price only recently more and more

elderly homes are including it in their buildings. One

example is the Besta-flats in Norway. But domotics,

as some robots, are made for a broader public and

only later repurposed for the elderly. However they

also offer big advantages: they are passive and non-

intrusive technologies which make them very

comfortable and safe for the elderly.

iTV (interactive Television) was originally

represented as a means of linking individuals

together by providing each with an electronically

mediated representation of the other’s voice and

visual presence (Wellens, 1979), providing a way of

deploying video telephony service. This type of user

interaction based on “social television” concept, is

nowadays being exploited into many other different

new services for all the inhabitants at home, thanks

to the development of the Internet, broadband

communications and specific devices such as STBs

(Set-Top-Boxes) and mobile phones.

Obviously, four groups are complementary,

meaning that only a combination of technologies

could be positive and really profitable for the

elderly. Moreover, through these complementarities

and their own progresses, these technologies and

services for users at home seem to trend to a

seamless integration. For example, some domotic

systems include teleassistance service (PERS) -even

a step ahead with a telemedicine service-, and can be

controlled by a TV set with a remote control or even

a touch screen. Thus presented research, referred

here as AmIE (AmIE, 2008), tries to observe the

best practices from technologies discussed above,

emphasising unique features of non-intrusiveness

and versatile adaptability of the offered services.

3 AmIE SERVICE SYSTEM

The main building blocks of AmIE service system

are presented in Fig.1. The heart of the systems is

the Intelligent Rule Engine. The AmIE system

operation is based on the rules which are defined by

means of interactions with health care professionals.

Rules define when and how to provide help,

compensate some behaviours, uplift the person, keep

track of the behaviour and send alarm signals. The

system also monitors vital parameters such as

pulsioximetry, electrocardiography, temperature,

blood pressure, heart rhythm, breathing and glucose

level. In this case rules enable the system to

understand for example when those medical

parameters are low or high and consequently

generate an alarm, when and how the system should

interact with the user in case of behavioural

disturbances, emotional or memory problems and to

support the user during activities of daily living.

Figure 1: AmIE service system – building blocks.

An environment equipped with an infrastructure

of sensors and actuators, including body sensors,

medical devices, domotics equipment, RFID tags, as

well as sensors of various kinds (temperature,

humidity, pressure, movement and presence

detectors, etc.) monitors what happens in the space

and collects data of various kinds and log

behaviours. Data collected by the sensors are

processed, interpreted, stored, correlated, and

ultimately mapped into decisions and decision rules.

AmIE will be able to discover hidden information

applying Data Mining techniques over collected

data, identifying, in some cases, patterns and

correlations between these data that could reveal

trends and future users’ behaviour, making in this

HEALTHINF 2009 - International Conference on Health Informatics

422

way predictions based in the medical and

behavioural history of each user. Key components of

service system are context-awareness and databases

whose task is to digest all the raw data and to learn

and act based upon it. Such system intelligence is

partly centralized (i.e. home server) and partly

distributed.

The adaptive system intelligence enabled by

innovative developments and technologies in the

domain of sensors data fusion, semantic modelling

and automatic reasoning of information coming

from different sources (sensors, user profiles,

calendars, home agenda) will insure the reliable

personalised on demand service delivery.

Security and privacy implications are concerned

particularly when employing sensing technology and

recording/monitoring behaviour of people.

Furthermore, the issue of personalization involves

storing and using an embedded user model for the

purposes of automatically selecting and adapting

functionality and services offered. It is important to

keep the user informed about the data that will be

collected and how it will be used. Obviously, this is

all related to legal and ethical issues too, managed

by an Ethical Committee.

3.1 Applications and Services

The scenarios below give an idea about applications

and services provided by AmIE system.

Wellness Evaluation. Anna is a 75-year-old woman

who lives in her own apartment. AmIE system

gathers automatically online information about her

current wellness status. An online wellness profile is

estimated by combining data from several sources.

The system collects sensor data from wearable

sensors, environment sensors, wellness self-

evaluations, social proximity (nurse, family,

friends,…), health record information, etc. The

online profile can be utilized by Anna herself,

relatives and health care professionals to personalize

the services offered to Anna. Anna has several

wearable sensors, such as a motion detector in her

wrist watch. The system notices when Anna is active

and performing her exercises, and also when Anna is

passive and not doing much. The AmIE system

stores also some information from bed sensors about

the sleep quality, about how many times certain

electrical appliances are used or different doors

opened during the day, how many phone calls or

visitors Anna has had today, etc. Anna’s mental

wellbeing is analysed through interactive games,

such as memory games. In addition, information

from possible medical measurement devices, such as

blood pressure meter, is collected by the system.

Anna can give her own assessment on her current

condition to the system by touching a corresponding

smiley face. A nurse or a family member visiting

Anna can give their opinion on Anna’s condition

using the same method or by using a web or mobile

phone service. This information together with sensor

information described above is combined to indicate

Anna’s current condition with e.g. “traffic light

indicator”.

Supporting Independent Living. Dealing with the

activities of the daily living, Anna is assisted in

doing the laundry, cooking and shopping. A washing

machine equipped with RFID tags may warn about

incompatibilities among the clothes being entered

into the machine. A refrigerator keeping track of the

foodstuff available may refresh the shopping list or

advice some alternative food recipes based on user

preferences, diets or health recommendations. Using

a simple device with several buttons for

communication, Anna may contact a call center for

further assistance in reserving time to barber or ask

for any other home assistance like cleaning or

shopping. Easy to use video TV communication is

available in Anna’s home to communicate with

friends, family, or share photos and gaming, but also

to communicate with medical staff if she feels some

need for this.

4 MARKET OPPORTUNITIES

Nowadays, various systems for home assistance are

already being commercialised. Companies such as

Tunstall, Mextal, Philips and Siemens

Communications and Tadiran Spectralink Ltd in

Europe or Telehealth Broadband in the US,

commercialise teleassistance and telemedicine

systems.

AmIE offers several innovations that constitute

strengths also in the market. AmIE does not only

perform a diagnosis but also predicts future

problems and helps the user improve his/her daily

habits to remain healthy for a longer time. The

system is expected to have a highly degree of

acceptance by the elderly users since the interaction

is not a roughly configured one, but adapted to each

user and situation (multimodal interaction).

Moreover, its design is based on the principle of

non-intrusiveness so that the user’s daily life is not

affected by the system elements. AmIE is positioned

at the boundary between self-monitoring and life-

style applications on one side and professional

monitoring and elderly care at the other side (see

Fig.2). This positioning has the advantage of

evolving along with the user from comfort services

and life style applications to assisted living in a

natural way.

AmIE - Towards Ambient Intelligence for the Ageing Citizens

423

Figure 2: The position of AmIE in the market. Source:

Parks Associates 2008.

The above discussions show that there is a good

possibility to address market and add-value service

delivery for elder. However, there are some

substantial challenges that still have to be dealt with:

 Health is such a delicate matter, that it can not be

left in an automatic system’s hands.

 Skills to interact with a computer: the elderly are

quite reluctant to technology and are not usually

accustomed to interacting with it.

 Who is our customer? It is found that in different

areas Age Care is organized with different model

(Scandinavia vs. South Europe). Payer and end-

user is not always the same person.

 Do we know the real needs/resources of the

elderly to solve our task in a proper way? Old

people are not just one target group based on age,

instead a huge group of individuals with different

needs and ever changing mental, physical and

perceptual capabilities.

5 CONCLUSIONS

In comparison to other available solutions, AmIE

will act as a preventing and assisting intelligent

system. Still several R&D challenges need to be

addressed to achieve its ambitious goals:

Reasoning and Intelligence. In order to add

intelligence to the system, technologies related with

Artificial Intelligence are applied. However there is

still a challenge to build a reliable usage and context

models from low-level sensors. Thus the health care

professional help will be involved to ensure that

collected data covers wide range of situations.

Living Ontologies. Ontologies are indispensable

elements of the AmIE rule definition module.

However grounding of ontologies in evolving, real-

world communities and service activation process is

still an open issue.

User Information. One key issue for AmIE is the

acquisition of information about the user, its

processing and analysis to derive predictions and

intelligent interaction with the user. Personal up-to-

date data access is a risen challenge concerned with

architectural, regulator and commercial issues.

REFERENCES

Shinozaki, R., at al., 2005. Communication and Control of

a Home Robot Using a Mobile Phone. 6th Pacific Rim

Conference on Multimedia, Jeju Island, Korea,

November 13-16

Porteus, J. and Brownsell, S., 2000. Using Telecare.

Exploring Technologies for Independent Living for

Older People. Brighton: Anchor Trust, Pavilion.

Publishing

Kenchiku, K., 2005. Introducing New PC emergency Call

Systems. Presentation at the IAHSA Symposium in

Norway, June 27-29.

Fellbaum, K. and M. Hampicke, 2002. Human-Computer

Interaction in a Smart Home Environment. Symposium

"Domotics and Networking", Miami, Nov 11-12.

Van Berlo, A., 2005. Smart Houses and smart living for

senior citizens: chances and markets. Silver Economy

in Europe Conference in Bonn, Germany Feb 16-17.

Van Berlo, A. and Fellbaum, K., 1999. Smart Home

Technology: Useful Applications for older people.

Assistive technology Research Series 6, pp.508-512.

Amsterdam.

MOBILALARM, 2005. Validating European Mobile

Alarm Services for Inclusion and Independent Living.

D 2.1 Market Analysis.

Finkelstein, S.M. Speedie, S., Potthoff, S., and Ratner, E.,

2005. Home Telehealth Provides an Independent

Living Option for the Elderly, In proceedings

TELEHEALTH 2005 Banff Canada July 17-19

Eurostat Publications Office, 2008. EUROPE IN

FIGURES, Eurostat Yearbook 2008

AmIE project home, http://www.amieproject.com/, 2008

HEALTHINF 2009 - International Conference on Health Informatics

424