Challenges for the Realization of Ambient Intelligence

(Extended Abstract to the Keynote on)

Juan Carlos Augusto

School of Computing and Mathematics and CSRI, University of Ulster, UK

jc.augusto@ulster.ac.uk

Abstract. A decade after the ISTAG report, Ambient Intelligence is still not a
concept which is widespread in industry and even less in society. This is not a
surprise, nobody working in this area claimed it will be a trivial process or that
can be achieved in a couple of years. There are still formidable challenges to be
tackled by the scientific community and there is no guarantee we will solve all
of them, but the more progress we make on some of those the closer we will be
from a stage where some of the initial aims are adopted. This talk will highlight
those challenges, explain why they are important and put forward the idea that
we can possibly do significant progress exploring those avenues for innovation.

Keywords: Ambient Intelligence, reliability, interaction, privacy.

1 Introduction

For centuries humans have witnessed scientific and technological leaps that changed
the lives of their generation, and those to come, forever. We are no exception. In fact,
so many of those advances are occurring now, in a more or less unperceived way.
Slowly and silently technology is becoming interwoven in our lives in the form of a
variety of intelligent devices which are starting to be used by people of all ages.
Advances in miniaturization of microprocessors have made possible a significant
development for Ambient Intelligence. Computing power is now embedded in many
everyday objects like home appliances (e.g., programmable washing machines, mi-
crowave ovens, robotic hovering machines, and robotic mowers). Intelligence travels
with us outside the home (e.g., mobile phones and PDAs), and they help guide us to
and from our home, safely and economically (e.g., car suspension and fuel consump-
tion and GPS navigation). The greening of Information Technology has delivered
computers that require reduced power, virtualization of servers, cloud computing and
network protocols that actively manage power to conserve energy (e.g. Bluetooth
4.0). These technological developments have merged with advances in the areas of
ubiquitous and pervasive computing which in turn were complemented by other exist-
ing areas of computing (for example artificial intelligence, HCI, and networking) into
one new and exciting are which aims at providing digital environments which support
people in their daily lives by assisting them in a ‘sensible’ way [1].

James J. (Jong Hyuk) Park et al. (eds.), Human Centric Technology and Service in Smart Space,
LNEE 182, pp. 65–69, DOI: 10.1007/978-94-007-5085-2_9,
© Springer Science+Business Media Dordrecht 2012
66 J.C. Augusto

Having the necessary technology, however, is not enough for an area of science to
flourish. Experiences of people with computers over recent decades have created an
interesting context where people’s expectations of these systems are growing and
their fear of using them has decreased. Concomitantly with this difference in the way
society perceives technology there is also a change in the way services are handled.
Technology is being adopted by all generations, now including older people,
exemplified by the ‘baby boomers’ who have begun to embrace new technological
advances.
More than a decade after the European Commission took a visionary step in
requesting analysis of the potential of Ambient Intelligence [2]. In the intervening
decade the area of Ambient Intelligence and related concepts like Smart/Intelligent
Environments have been explored to assess the extent they can be successfully de-
ployed and integrated into our society [3–6]. The realization of Ambient Intelligence
requires the engineering of technological solutions which are helpful to people,
transparent, reliable and affordable.
This article provides an overview of the Keynote offered at HumanCom2012
which is focused on a critical analysis of the area focusing on potential barriers for the
realization of Ambient Intelligence. We believe this will help to focus efforts on the
elements that will have a higher impact on realizing the concept of Ambient Intelli-
gence into ‘real, deployed’ applications which change positively the way people
interact with technology in their daily lives.

2 Inherent Challenges

The very nature of the area means we need to accept a number of serious challenges
which make our work at the same time difficult and interesting. This section lists
some of the most important and core challenges.

2.1 Users

The areas which have developed under the names of ubiquitous computing, pervasive
computing, people centred computing, ambient intelligence, smart environments and
Intelligent Environments are all closely related, arguably with more in common than
differences. One of those common aims of all these areas is that one of putting
humans at the very centre of the system and on empowering them:

“…the human is the master and the computer the slave and not the other
way round.” [7].

These systems should be able to help people of all ages and educational backgrounds
(crucially those who do not have IT knowledge). So many of the systems built on labs
today are more for the personal gratification of the ego of the researchers than to pro-
vide real support to real people or to answer a request from users. Unobtrusiveness
and transparency of services were advocated from the very beginning [8] as core
 Challenges for the Realization of Ambient Intelligence 67

concepts to be respected. Still, so often these features are forgotten and other aims
become more relevant, e.g., having a system ‘working’.
System-Oriented (Importunate) Smartness is more often achieved but it has been
correctly advocated that People-Oriented (Empowering) Smartness [9] is rather what
is needed.
Systems also have to be resilient enough to cope with: users which will try to use
the system in unexpected ways and with the richness and variability of human’s
behaviour on a daily basis. [10]

2.2 Environments

We can define an environment as “the area (physical space) that the sensors can
sense”. One classification of environments is in:
─ closed spaces with relatively well defined boundaries (e.g., houses, hospitals, class-
rooms, and cars)
─ open spaces which do not have well defined boundaries (e.g. streets, bridges and
car parks, fields, and sea).

They are usually rich, complex, unpredictable, possibly generating substantial ’noisy’
data, unstructured and sometimes highly dynamic.

2.3 Perception

All these systems are embedded in a world they have to act upon. Sensors allow the
system to perceive what happens in a place without a human being necessarily being
there. Perception is as good or bad as the sensing network. This sensing network in-
forms but at the same time oversimplifies reality.

3 Current Limitations

Inherent Challenges cannot be solved, they are part of the rules of the game we accept
to play. Current limitations are problems we can realistically aspire to work on
(achieve better results and hence increase user acceptance). Next we list some of the
problems awaiting solution.

3.1 Accurate Context-Awareness

Intelligent environments deal with massive amounts of data and highly complex situa-
tions. As a result, they also need to make decisions based on insufficient, incomplete
and noisy data samples.

3.2 Preferences vs. Needs

Preferences and needs distinguish us from one another. We change them often due to
unforeseen circumstances or to our own decision. It will be unpractical to design each
68 J.C. Augusto

system from scratch in an ad-hoc manner for each person but a “one size fits all” ap-
proach will be unacceptable.

3.3 ‘Mindreading’

Some users may not be willing to speak or to use any keyboard or device to explicitly
indicate how they feel. Is there any way a system can understand whether the user is
in more of a receptive or introverted mood?

3.4 Coping with Multiple Occupancy

Identifying users in an unobtrusive, reliable and affordable way is still a challenge

for all systems. Following users as they move is a further complication. Delivering
services in shared spaces where there are conflicting preferences/needs, e.g., air
conditioned in a car, T.V. settings in a living room, message boards in public
spaces.

3.5 Deploying Reliable Systems

Systems fail. The more complex, the more often. Systems grow in complexity and
responsibility. Still, humans tend to trust systems after seeing them working well for a
while more consideration on what to do when systems fail to assist and people is left
alone.

3.6 Ensuring Ethics and Privacy

Many of these systems which are human-centred can collect fantastic amounts of data
from users and their lifestyles. How to make sure this stays private and is not
misused?

4 Conclusions and Reflection

Applications of pervasive and embedded computing have been an important driving
force on the materialization of a society supported by information. Such applications
are at the core of fundamental areas such as healthcare and transportation. As a rela-
tively new and developing area there are still important barriers for optimal develop-
ment and widespread acceptance and adoption. The problems are big enough to re-
quire more community approach (instead of isolated efforts).

References
1. Augusto, J.C.: Ambient Intelligence: the Confluence of Ubiquitous/Pervasive Computing
and Artificial Intelligence. In: Intelligent Computing Everywhere, pp. 213–234. Springer
(2007)
2. Ducatel, K., Bogdanowicz, M., Scapolo, F., Leijten, J., Burgelman, J.-C.: Scenarios for
Ambient Intelligence in 2010. IPTS, Seville (2001)
 Challenges for the Realization of Ambient Intelligence 69

3. Aarts, E., Harwig, R., Schuurmans, M.: Ambient Intelligence. In: Denning, P. (ed.) The
Invisible Future, pp. 235–250. McGraw Hill, New York (2001)
4. Aarts, E.: Ambient Intelligence: A Multimedia Perspective. IEEE Multi Media 11(1),
12–19 (2004)
5. Cook, D.J., Augusto, J.C., Jakkula, V.R.: Ambient Intelligence: applications in society and
opportunities for AI. Pervasive and Mobile Computing 5, 277–298 (2009)
6. Nakashima, H., Aghajan, H., Augusto, J.C. (eds.): Handbook on Ambient Intelligence and
Smart Environments. Springer (2009)
7. Dertouzos, M.: Human-centered Systems. In: Denning (ed.) The Invisible Future, pp. 181–
192
8. Weiser, M.: The computer for the 21st century. Scientific American 265(3), 94–104 (1991)
9. Yuan, F., Prante, T., Röcker, C., van Alphen, D., Stenzel, R., Magerkurth, C., Lahlou, S.,
Nosulenko, V., Jegou, F., Sonder, F., Plewe, D.: Smart Artefacts as Affordances for
Awareness in Distributed Teams. In: Yuan, F., Kameas, A.D., Mavrommati, I. (eds.) The
Disappearing Computer. LNCS, vol. 4500, pp. 3–29. Springer, Heidelberg (2007)
10. Satoh, I.: An Agent-Based Framework for Context-Aware Digital Signage. In: Proceedings
of ISAmI 2010, Salamanca, Spain, pp. 105–112 (2010)