Reflecting on the Ambient Intelligence Vision
A Cyber-Physical-Social Perspective
Olga Murdoch, Michael O’Grady, Rem Collier and Gregory M. P. O’Hare
CLARITY Centre for Sensor Web Technologies, University College Dublin, Belfield, Dublin 4, Ireland
olga.murdoch@ucdconnect.ie, michael.j.ogrady@ucd.ie, rem.collier@ucd.ie, gregory.ohare@ucd.ie
Keywords:
Ambient Intelligence, Cyber Physical Social Systems, Middleware, Web Technologies, Ubiquitous Comput-
ing.
Abstract:
By acquiring and reasoning about user and environmental context, Ambient Intelligence (AmI) systems enable
intelligent and intuitive interactions between people and their physical environments. AmI traditionally seeks
to build on ubiquitous sensing technologies and communications to acheive this aim. Over a decade later,
people are frequently immersed in or engaged with alternative, software, environments, such as those provided
as web applications or social networks. Now, to form more complete contextual representations of users, AmI
solutions require the context of multiple environments in which users may be immersed. This paper proposes a
middleware supported framework for cyber-physical-social AmI that relies on the combined efforts of modern
web technologies and ubiquitous computing as the enabling technologies.
1 INTRODUCTION
In a modern data-saturated world, knowledge and
inference-driven systems integrate intelligence into
real-time services and devices allowing for the pro-
vision of diverse, self-adaptive, autonomous, person-
alised and intelligent computing applications (Weiser,
1991). This is the vision of ubiquitous computing
research which aims to provide unobtrusive deploy-
ments of technology enabling the provision of Ambi-
ent Intelligent (AmI) solutions. AmI (Sadri, 2011) en-
ables autonomous environments and, specifically, in-
tuitive interaction through the provision of Intelligent
User Interfaces (IUIs) enabled by in-situ intelligence
and embedded decision making.
While Wireless Sensor Network (WSN) and asso-
ciated middleware research focuses on enabling this
goal, the World Wide Web (WWW) has become a
modern day source of personal and informative, real
time and archival data. Through user participation in
social networks and use of personalised web services,
the social web (Chi, 2008) has become an extension
of people’s everyday lives. Social and non-social web
sites provide real time reports of topical information,
breaking news and environmental information such as
weather reports and regional disaster alerts. Further-
more, virtual and augmented reality research is be-
ginning to enable real-time interactions with environ-
ments that are very different to the physical world we
live in.
Acknowledging these advances in technology re-
quires us to extend our understanding of context, envi-
ronments and sensors. The enabling technologies for
AmI must now accommodate modern environments
and leverage advances in WWW research. Further,
the implications of bridging the cyber-physical-social
divide must be assessed so as not to limit user adop-
tion of cyber-physical-social AmI solutions.
This paper summarises advances in Ubiquitous
Computing and the evolution of the World Wide
Web (Section 2). We then reflect on the traditional
AmI vision (Section 3), offering a current perspec-
tive on what we consider context, environments and
sensors, and addressing the benefits and challenges
of cyber-physical-social AmI. This analysis motivates
our proposed middleware-supported framework for
cyber-physical-social AmI, which composes existing
research efforts into a unified architecture (Section
4). Conclusions and a research agenda are offered to
deliver the foundations of cyber-physical-social AmI
(Section 5).
2 THE WORLD WIDE WEB AND
UBIQUITOUS COMPUTING
The World Wide Web and Ubiquitous Computing
are diverse yet converging areas of research, both of
229
Murdoch O., O’Grady M., Collier R. and O’Hare G.
Reflecting on the Ambient Intelligence VisionA Cyber-Physical-Social Perspective.
DOI: 10.5220/0005426502290235
In Proceedings of the Fourth International Symposium on Business Modeling and Software Design (BMSD 2014), pages 229-235
ISBN: 978-989-758-032-1
Copyright
c
2014 by SCITEPRESS – Science and Technology Publications, Lda. All rights reserved
which strive to produce personalised and intelligent
solutions that will be adopted by and meet the needs
of users. This section provides an overview of state
of the art research in these areas followed by analysis
of the supporting frameworks that contribute to our
unified research agenda.
2.1 World Wide Web
Since the birth of the World Wide Web (WWW) in
1990 it has been transformed from a web of docu-
ments to a web of interconnected information, users
and services (Web 2.0). The emergence and wide
spread adoption of social networks in particular has
embedded users into the web by providing a mech-
anism through which they can connect and share.
By providing programmatic access to data through
standardised interfaces, this social web facilitates the
web of services, enabling enhanced and personalised
applications which further enhancing a user’s web-
based experience.
Figure 1: Semantic Web 3.0. (Hendler, 2009).
In 2009, Hendler proposed that Web 3.0 will be
in part the realisation of the semantic web (Hendler,
2009) as outlined by Tim Berner’s Lee in 2001
(Berners-Lee et al., 2001). It is stated that Web 3.0
can be viewed as “Semantic Web technologies in-
tegrated into, or powering, large-scale Web applica-
tions”. This is illustrated in Figure 1 which identifies
the underlying semantic technologies that link Web
2.0 data allowing the formation of a Semantic Web
upon which Web 3.0 applications will strive. Web 3.0
applications differ from those we know to be Web 2.0
in that they will become ‘smarter’ due to the ease of
performing simple inferences with the support of se-
mantic knowledge bases.
In 2010, Hendler and Berners-Lee (Hendler and
Berners-Lee, 2010) argued that the social connected
nature of current web systems is an early realisation
of web connected social machines that will achieve
much more through collaborative functionalities. It
is argued that the future web must “be designed to
allow the virtually unlimited interaction of the Web
of people”. It is envisioned that people will build and
share their own social machines.
While social machines have not yet been realised,
Web 2.0 persists in the mainstream and Web 3.0 appli-
cations are beginning to emerge in the research litera-
ture (Hendler, 2009). Initial insights into Web 4.0 and
beyond are now offered in research literature. Aghaei
(Aghaei et al., 2012) and Kambil (Kambil, 2008)
speculate, independently, on what is to come after
Web 3.0, agreeing that Web 4.0 will be a realisation of
the WWW where intelligent interactions between hu-
mans and machines take place. Kambil offers further
insights into Web 5.0, proposing a sensory-emotive
web, a WWW that is responsive to users emotional
state or receptivity to specific information.
2.2 Ubiquitous Computing
Ubiquitous computing involves building highly re-
sponsive and self-adaptive systems which aim to in-
crease the integration of technology into the fabric of
everyday living (Weiser, 1991), for example, smart
spaces and environmental monitoring. Enabling this
vision requires the availability and use of heteroge-
neous sensing devices that enable data to be captured
in real time. AmI by its very nature requires that sen-
sors and services be unobtrusive to a user, necessitat-
ing the provision of sensors that can be embedded in
the environment and existing user devices is essential.
Smart spaces aim to provide ubiquitous (hidden)
services such as ambient personalised displays, mood
lighting, and automated services by embedding intel-
ligence in devices within a utilised space such as a
home or office. Internet of Things (IoT) and Wire-
less Sensor Network (WSN) researchers aim to enable
smart spaces through small to medium size deploy-
ments of smart things and sensors within an occupied
space.
Environmental monitoring systems typically
gather a diverse range of data relating to the condi-
tions of an environment enabling the development
of intelligent ecosystems that react to diverse phe-
nomena such as pollution, wildlife migration, natural
disaster or intrusion. Such systems require large scale
WSNs that can be left in the environment unattended
for long periods of time, meaning the provision of
sensing devices capable of intelligence is required.
The solution offered by the Sensor Web is “web
accessible sensor networks and archived sensor data
that can be accessed and discovered using standard
protocols and Application Programming Interfaces
(APIs)” (Botts et al., 2008).
Internet of Things (IoT), Web of Things (WoT)
and Sensor Web are terms often given conflicting def-
initions. Some IoT definitions (Atzori et al., 2010)
focus on internet enabled devices, while others em-
phasise the networking capability of such devices, not
Fourth International Symposium on Business Modeling and Software Design
230
unlike the WSN’s that underpin Sensor Web research.
This seems a natural progression from the IoT vision
that started in the RFID community where internet
enabled devices are limited in computational capabil-
ities. IoT researchers have since begun experiment-
ing with more powerful devices or smart objects that
are capable of inter-device communication amongst
other capabilities such as semantic reasoning and ac-
tuation. Web of Things (WoT) research, which is
seen as the next stage of the IoT, focusses on use of
emerging Web Standards to facilitate reuse of inter-
net enabled devices. The Sensor Web (Delin, 2002)
paradigm envisions connectivity of entities and users
in real and web-based environments through the adop-
tion of large scale heterogeneous sensor-actuator net-
works.
Such networks support intelligent sensing and ac-
tuating sometimes through the embedding of agents
(Tynan et al., 2005) sometimes even advocating Auto-
nomic Wireless Sensor Networks (Marsh et al., 2004).
They treat all entities as equal citizens regardless of
their origin, form or nature (O’Hare et al., 2012).
While IoT enabled applications described in research
tend to be ‘local’, such as smart spaces, Sensor Web
researchers strive to solve ‘global’ problems such as
environmental monitoring and early warning systems.
2.3 Supporting Frameworks
Middleware and programming frameworks aim to
ease the process of developing and deploying systems
that typically compete for resources and are depen-
dent on complex underlying technologies. While the
two terms are often used interchangably within re-
search literature, here we provide a distinction. Mid-
dleware frameworks ease and manage the deployment
of underlying technologies that enable data access and
management, and also facilitate application deploy-
ment and access to resources. Programming frame-
works enable rapid development of user-facing proto-
types and applications through the provision of API’s
that enable abstracted access to the data managed
by the middleware. An end-to-end support frame-
work comprises both middleware and programming
technologies. Each of WWW, IoT, WSN, and Sen-
sorWeb research domains employ such supporting
frameworks to allow developers of user facing solu-
tions to remain application domain experts, without
concern for the underlying infrastructure upon which
their application runs.
Given the nature of the current WWW, which can
be described as Web 2.0, a Social Web, and a Web of
Services, Service Oriented Architectures (SOA) have
become widely adopted as the supporting framework
for Web application developers (Issarny et al., 2011).
Service Oriented Middleware for the WWW usually
focuses on service discovery and composition (Mi-
lanovic and Malek, 2004) while programming sup-
port is typically provided through mashup tools that
allow developers use mutliple diverse services within
a single application through an easy to use interface
(Grammel and Storey, 2010). Cloud Computing in-
frastructures have emerged as a powerful tool for sup-
porting rapid deployment of services and applications
(Armbrust et al., 2010).
Middleware and programming frameworks for
ubiquitous computing aim to ease the process of de-
ploying, potentially large scale, WSNs, enabling de-
velopers to build data-driven solutions without the
learning curve involved with network deployment
and programming. There is a wealth of middle-
ware research in the broad ubiquitous computing and
WSN domains (Raychoudhury et al., 2013) (Wang
et al., 2008) (Hadim and Nader, 2006), much of
which provides solutions to distinct problems in spe-
cific domains, or only supports the use of homoge-
nous sensors. A generalised middleware for inher-
ently heterogeneous environments was called for in
2001 (Geihs, 2001); over a decade later, it has still
not been realised. Due to the flexibility it offers in
terms of reusability, extensibility and interoperabil-
ity of services, SOA is becomming a standard ap-
proach for developing middleware solutions for the
Internet of Things, WSN and Sensor Web research
(Mohamed and Al-Jaroodi, 2011) (Chu and Buyya,
2007). SOA partially overcomes the heterogeinity
problem bringing middleware researchers closer to re-
alising a generic middleware solution for diverse sys-
tems.
3 REFLECTING ON THE AmI
VISION
Many of the developments described in the previous
sections have occurred within the last decade. In
the case of AmI, the question arises as to what are
the implications for paradigm itself and how or if it
should reinvent itself in light of these developments.
While the challenges of WWW and ubiquitous com-
puting research differ, research and developments into
the supporting frameworks for such systems are con-
verging towards common, SOA, solutions and a sin-
gle unifying framework is expected to emerge. The
divide between WWW (cyber-social) and ubiquitous
computing (physical-social) will be bridged, enabling
a platform for cyber-physical-social AmI solutions.
Bridging the cyber-physical-social divide is of benefit
Reflecting on the Ambient Intelligence Vision - A Cyber-Physical-Social Perspective
231
to the AmI paradigm. How it is accomplished must be
accommodating of the users that will adopt AmI tech-
nologies and are core to the success of AmI research.
Here we discuss cyber-physical-social context, envi-
ronments and sensors and explore both the benefits
and challenges of bridging the cyber-physical-social
divide within the context of AmI research.
3.1 Cyber-Physical-Social Context,
Environments and Sensors
Context comprises any information that character-
izes the situation of any person, place, or object that
is considered relevant to the interaction between a
user and an application (Abowd et al., 1999). For
AmI, context is typically aquired through deploy-
ments of ubiquitous sensing technologies monitoring
users, their immediate physical environment and the
things they interact with. In keeping with this defini-
tion of context, cyber-physical-social AmI requires a
broader understanding of environments and sensors.
An environment can be considered to be the geo-
graphical location and immediate surroundings within
which a user or entity currently resides or equally the
cyber (web-based/software) environment that a user
or entity is engaged with or immersed in. Modern
AmI users are increasingly likely to be immersed in
environments alternative to the one they are physi-
cally present in, such as social networks, games, e-
commerce, and virtual/augmented reality. This must
be accommodated by a sensing infrastructure en-
abling cyber-physical-social AmI.
While sensors are typically defined by the features
of the device that enable it to sense, here we refocus
on the capabilities regardless of the form of the sen-
sor. A cyber sensor is a software sensor that monitors
any programmatically accessible environment, user or
entity (O’Grady et al., 2013). A sensor that is intel-
ligent, as distinct from operating continuously in a
sense-transmit cycle, can monitor its own state and
that of connected sensors. It can also decide when
it is appropriate to report a phenomenon, reducing
communication cost which can have a significant im-
pact on the operational lifespan of a battery-powered
sensor, or reducing traffic to a server which is being
queried by a cyber sensor. A supporting infrastructure
for cyber-physical-social AmI must treat cyber sen-
sors and physical sensing devices as equal citizens.
3.2 Cyber-Physical-Social AmI:
Challenges and Benefits
A core challenge for AmI has traditionally been user
adoption of AmI solutions. The importance of intu-
itive interaction is as important now as it was when
Weiser proposed the original ubiquitous computing
vision, yet remains a distant goal. In the case of
AmI, if the emphasis on the physical element, that
is ubiquitous computing, is augmented with a social
perspective, this may inform the behaviour of AmI
systems and make them more human-centric. Cyber-
physical-social AmI solutions will be based on a more
complete contextual foundation of its users than the
knowledge base achievable through embedded sen-
sors alone. Difficulties in assessing a user’s mood or
motivations will be addressed through the inclusion
of a user’s real time social web presence within the
solution.
The traditional vision of calm computing encom-
passes technologies that stay hidden and make deci-
sions without any user input. Motivated by human
ability to exploit their environments, Rogers (Rogers,
2006) describes an alternative goal to Weiser’s calm
computing that is engaging computing, shifting from
‘proactive computing’ to supporting ‘proactive peo-
ple’. It is also argued that we “simply don’t do ‘smart’
well yet” (Greenfield, 2006), but perhaps that is be-
cause we don’t have access to comprehensive histor-
ical data for individual users such that a system may
learn to adapt appropriately. It is proposed here that
the provision of engaging technologies is a necessary
step in the journey to ultimately providing calm com-
puting. As users become more engaged with (and
accepting of) technology, they will learn to under-
stand and trust in technological advances. Simulta-
neously, researchers will learn what users are will-
ing to accept from technology, both in terms of what
tasks/decisions they are willing to hand over to auto-
mated solutions and what level of control users de-
mand over their own data. This intermediate step
between engaging and calm computing will facili-
tate the organic growth of large user annotated data
sets (aggregated through user interactions with tech-
nology), providing a new historical knowledge base
upon which local, personalised, intelligent solutions
can become automated to suit AmI users.
While it is envisaged that middleware will bridge
the cyber-physical-social divide for AmI solutions, it
must be considered how that middleware will oper-
ate. While services exist that facilitate the sharing of
WSN, or Internet of Things, data to the web, the pub-
lication of private sensor data to an external data bro-
ker will lead to concerns surrounding user privacy and
ownership of data. The idea that all user information,
including that originating from sensors embedded in
private environments, is managed by third party data
brokers goes against modern social expectations and
will lead to resistance in terms of adopting these new
Fourth International Symposium on Business Modeling and Software Design
232
technologies. Advances in information retrieval, dis-
semination and capture technologies far exceed those
in areas of web data providence and privacy. A con-
tributor of information to the current WWW must un-
derstand the privacy controls of hugely diverse service
providers whose terms and conditions may change at
any point. Relying on these external data brokers
to control and secure information means users can-
not be certain of who has access to their information
and for what purpose. It is therefore necessary when
bridging the cyber-physical-social divide, to provide
a safe place where information processing and intel-
ligent decision making can take place while adhering
to user defined privacy requirements. This results in
the requirement for web-based data to be integrated
with WSN data within a local distributed middleware
that supports fully customisable privacy and security
solutions. Such an approach will facilitate fusion of
widely diverse user data, integrating WWW, Sensor
Web and IoT research efforts to date, as proposed in
the following section.
4 PROPOSED
CYBER-PHYSICAL-SOCIAL
ENABLEMENT FOR AmI
In Section 2, Service Oriented Architecture (SOA)
was identified as a common approach to enabling
WWW, IoT and Sensor Web solutions. Despite the
differences between WWW and ubiquitous comput-
ing, a mapping can be made between the challenges to
be addressed by services, enabling a unifying solution
to bridge the cyber-physical-social divide. An adapta-
tion of a layered SOA architecture for cyber-physical-
social enablement is described as follows (adapted
from (Atzori et al., 2010)):
Objects. Sensors, Smart devices, WSNs, Targeted
aspects of any software including WWW sites and
services
Object Abstraction. Software the unifies access to
and control of objects regardless of original form,
converting sensed data to a standardised format
Service Management. Software services that pro-
vides features to simplify deployment and man-
agement of a complete infrastructure, e.g., stan-
dards, object discovery, configuration and status
monitoring. Trust, privacy and security, if not in-
built to the core of the middleware architecure, are
typically provided as services that can be config-
ured as required
Service Composition. Supports the application de-
veloper in composing applications supported by
the infrastructure through the composition of pre-
defined services exposed through an API
Applications. Multiple, concurrent appplications
competing for infrastructure resources in a way
that is managed by the previous layers according
to its configuration
A supporting framework for cyber-physical-social
AmI will accommodate semantic reasoning and di-
verse artificial intelligence solutions (Murdoch and
Nixon, 2010) enabling the development of engaging
AmI solutions through which user feedback will sup-
port enhancement of automated services and more in-
tuitive interactions.
Figure 2: (A) Open Web 3.0 and Web of Things (B) Private
Webs of People co-existing with Open Web 3.0.
To further understand the requirements of an en-
abling framework for cyber-physical-social AmI, we
consider the wider perspective, how IoT, Sensor Web
and WWW research agendas will unify to acheive this
goal. In Section 2 we identified current and projected
future trends in WWW research. In Figure 2 (A) we
extend Figure 1 to illustrate how the Web of Things
integrates physical sensor data in a manner that is sep-
arate to the user contributed information obtained via
Web 2.0. This separation of concerns is not a true rep-
resentation of the reality which involves interactions
between people and things resulting in production of
data. A Web of People enabled by social machines is
an innovative vision that has the potential to unlock a
realm of previously unimaginable personalised appli-
cations and services. However, the notion of an open
social web and the current architecture of web tech-
nologies is in conflict with the necessity to provide
user friendly yet highly granular customisable privacy
controls.
The open WWW will only contain information
that is contributed with the intention of being open
Reflecting on the Ambient Intelligence Vision - A Cyber-Physical-Social Perspective
233
and reused. We propose that the answer is not exclu-
sively in finding better privacy solutions but in under-
standing how, why and when to use the open WWW.
Then, we need to facilitate the evolution of personal
social machines that are created from the ground up.
People will be empowered to be their own data broker,
privacy rules for individual pieces of data will travel
with that data as it is shared/disseminated through net-
works of users. Data that is not deemed public will
not be shared to the open web. In Figure 2 (B) we
extend the Web 3.0 framework to allow for personal
sensor webs, enabling a web of people. A sensor web
is a collection of sensors, in this case, monitoring all
entities relating to a particular user and their environ-
ment. A private sensor web comprises a user, their
devices, their personal environments, and information
about their connections and connected environments
to which the users have been granted mutual access.
It is proposed that a private sensor web will be the en-
abling technology for personal social machines with
fine grained user control over privacy and organically
grown connections between users supported by dis-
tributed mobile social networks. The separation of
concerns between private user information and that
shared publicly via Web 2.0 and the Web of Things
will allow for bespoke privacy mechanisms that are
user controlled, while the open web remains unthreat-
ened.
Figure 3: Middleware Enablement for Cyber-Physical-
Social Web Greater Than 3.0.
Figure 3 further illustrates the proposed composi-
tion of ubiquitous and web technologies that will pro-
vide a foundation for future AmI solutions. Require-
ments for middleware and programming support are
highlighted throughout as is a layer depicting user-
centered, personalised, AmI solutions and interfaces.
It is envisaged that private sensor webs with connect-
ing users will form organically grown communities,
or webs of sensor webs. Privately sensed information
from a user’s physical or web-based environments
as well from publicly accessible data from Web 3.0
drive user-focussed services that adapt autonomously
to user needs as they change over time. Private, com-
munity, regional and global sensor webs will under-
pin future smart spaces and environmental monitor-
ing systems. It is anticipated that a data marketplace
will facilitate users selling high level abstracted data
to companies and service providers. The difference in
this model is that the data resides with the user until
they actively choose to share it. This will enable the
provision of data providence visualisations that allow
a user to understand how, where, when and why their
information is being used, empowering them to make
safer choices in terms of sharing private information.
Within the context of this unified cyber-physical-
social web, middleware is required to bridge, seam-
lessly, existing and future infrastructures includ-
ing WWW and Ubiquitous Computing deployments.
Programming Support is required to unify access to
the underlying infrastructures simplifying the devel-
opment process for application domain experts.
5 CONCLUSIONS AND
RESEARCH AGENDA
In this position paper, we considered AmI as a Cyber-
Physical-Social paradigm. Having reflected on cur-
rent state of the art in ubiquitous computing and web
research, a future cyber-physical-social web was pro-
posed as the contextual foundation that will underpin
Cyber-Physical-Social AmI. The need for a generic
middleware and software framework to enable this
vision of AmI was articulated and Service Oriented
Architecture identified as the appropriate solution for
acheiving this aim. Based on this discussion, the fol-
lowing research priorities are identified:
A deeper understanding of the need for cyber, in
particular social, data within ambient intelligence
solutions;
A seamless and unified abstraction model of cyber
(including social) and physical environments;
Unified abstractions for middleware services sup-
porting both WWW and ubiquitous computing in-
tiatives
An intuitive and feature rich middleware and pro-
gramming framework for the rapid development
of prototypes and testbeds.
This research agenda depends on collaborative ef-
forts of researchers in the broad fields of WWW and
Ubiquitous Computing research to deliver an infras-
tructure for user-centered cyber-physical-social AmI.
Fourth International Symposium on Business Modeling and Software Design
234
ACKNOWLEDGEMENTS
This work is supported by Science Foundation Ireland
(SFI) under grant 07/CE/11147 and the EU FP COB-
WEB Project under grant no. 308513.
REFERENCES
Abowd, G. D., Dey, A. K., Brown, P. J., Davies, N., Smith,
M., and Steggles, P. (1999). Towards a better under-
standing of context and context-awareness. In Hand-
held and ubiquitous computing. Springer.
Aghaei, S., Nematbakhsh, M. A., and Farsani, H. K. (2012).
Evolution of the world wide web: From web 1.0 to
web 4.0. International Journal of Web & Semantic
Technology, 3(1).
Armbrust, M., Fox, A., Griffith, R., Joseph, A. D., Katz,
R., Konwinski, A., Lee, G., Patterson, D., Rabkin, A.,
Stoica, I., et al. (2010). A view of cloud computing.
In Communications of the ACM. ACM.
Atzori, L., Iera, A., and Morabito, G. (2010). The internet
of things: A survey. In Computer networks. Elsevier.
Berners-Lee, T., Hendler, J., and Lassila, O. (2001). The
semantic web. Scientific American, 284(5):34–43.
Botts, M., Percivall, G., Reed, C., and Davidson, J. (2008).
sensor web enablement: Overview and high level ar-
chitecture. In GeoSensor networks. Springer.
Chi, E. H. (2008). The social web: Research and opportu-
nities. IEEE Computer, 41(9):88–91.
Chu, X. and Buyya, R. (2007). Service oriented sensor web.
In Sensor Networks and Configuration. Springer.
Delin, K. A. (2002). The sensor web: A macro-instrument
for coordinated sensing. Sensors, 2(7):270–285.
Geihs, K. (2001). Middleware challenges ahead. IEEE com-
puter, 34(6):24–31.
Grammel, L. and Storey, M.-A. (2010). A survey of mashup
development environments. In The smart internet.
Springer.
Greenfield, A. (2006). Everyware: The Dawning Age of
Ubiquitous Computing. Peachpit Press.
Hadim, S. and Nader, M. (2006). Middleware: Middle-
ware challenges and approaches for wireless sensor
networks. Distributed Systems Online, 7(3):1–1.
Hendler, J. (2009). Web 3.0 emerging. IEEE Computer,
42(1):111–113.
Hendler, J. and Berners-Lee, T. (2010). From the semantic
web to social machines: A research challenge for ai
on the world wide web. Artif. Intell., 174(2):156–161.
Issarny, V., Georgantas, N., Hachem, S., and Zarras (2011).
Service-oriented middleware for the future internet:
state of the art and research directions. In Journal of
Internet Services and Applications. Springer.
Kambil, A. (2008). What is your web 5.0 strategy? In
Journal of business strategy. Emerald Group Publish-
ing Limited.
Marsh, D., Tynan, R., OKane, D., and P OHare,
G. M. (2004). Autonomic wireless sensor networks.
Engineering Applications of Artificial Intelligence,
17(7):741–748.
Milanovic, N. and Malek, M. (2004). Current solutions for
web service composition. In IEEE Internet Comput-
ing. IEEE Educational Activities Department.
Mohamed, N. and Al-Jaroodi, J. (2011). A survey on
service-oriented middleware for wireless sensor net-
works. Service Oriented Computing and Applications.
Murdoch, O. and Nixon, P. (2010). A unifying framework
for situation identification methodologies. MRC.
O’Grady, M. J., Murdoch, O., Kroon, B., Lillis, D., Carr, D.,
Collier, R. W., and O’Hare, G. M. (2013). Pervasive
sensing: Addressing the heterogeneity problem. In
Journal of Physics: Conference Series, volume 450,
page 012044. IOP Publishing.
O’Hare, G. M. P., Muldoon, C., O’Grady, M. J., Collier,
R. W., Murdoch, O., and Carr, D. (2012). Sensor Web
Interaction. International Journal on Artificial Intelli-
gence Tools, 21(02):1240006.
Raychoudhury, V., Cao, J., Kumar, M., and Zhang, D.
(2013). Middleware for pervasive computing: A sur-
vey. Pervasive and Mobile Computing, 9(2):177–200.
Rogers, Y. (2006). Moving on from weiser’s vision of calm
computing: Engaging ubicomp experiences. In Ubi-
Comp 2006: Ubiquitous Computing, pages 404–421.
Springer.
Sadri, F. (2011). Ambient intelligence: A survey. ACM
Computing Surveys (CSUR), 43(4):36.
Tynan, R., Marsh, D., O’kane, D., and O’Hare, G. M.
(2005). Intelligent agents for wireless sensor net-
works. In Proceedings of the fourth international joint
conference on Autonomous agents and multiagent sys-
tems, pages 1179–1180. ACM.
Wang, M.-M., Cao, J.-N., Li, J., and Dasi, S. K. (2008).
Middleware for wireless sensor networks: A sur-
vey. Journal of computer science and technology,
23(3):305–326.
Weiser, M. (1991). The computer for the 21st century. Sci-
entific American, 265(3):94–104.
Reflecting on the Ambient Intelligence Vision - A Cyber-Physical-Social Perspective
235