THIRD GENERATION OF COMPUTING-VISION OF THE FUTURE

Abstract

Ubiquitous computing is unusual amongst technological research arenas. Most areas of

computer science research, such as programming language implementation, distributed
operating system design, or denotational semantics, are defined largely by technical problems,
and driven by building upon and elaborating a body of past results. Ubiquitous computing, by
contrast, encompasses a wide range of disparate technological areas brought together by a
focus upon a common vision. It is driven, then, not so much by the problems of the past but by
the possibilities of the future. Ubiquitous computing’s vision, however, is over a decade old at
this point, and we now inhabit the future imagined by its pioneers. The future, though, may not
have worked out as the field collectively imagined. In this paper, we explore the vision that has
driven the ubiquitous computing research agenda and the contemporary practice that has
emerged. Drawing on cross-cultural investigations of technology adoption, we argue for
developing a ‘‘ubicomp of the present’’ which takes the messiness of everyday life as a central
theme.

Key terms: - Ubiquitous, Distributed, semantics, technological research, central theme.

I. Introduction technology. The task of making this context

information available to components in
1. Overview computer systems has become a prerequisite
to advancing human-computer interaction
Research in Ubiquitous Computing has processes in Ubiquitous Computing. Context
arrived at a crossroad: A point of awareness, or more specifically how to
convergence where a technology create applications that are context aware, is
proliferated environment meets with the a central issue to Ubiquitous Computing
ability of people to interact with, and make research. Such research raises questions on
use of, the possibilities that this technology context acquisition, context representation,
creates. Advances in the various fields of distribution and abstraction, as well as
technology allow us to create artefacts and programming paradigms, development
environments that provide computing and support, and implications on human-
communication resources. The computer interaction in general. The
understanding of how humans will interact research presented in this thesis concentrates
and make use of such systems is however on some of these issues. First the question of
largely unresolved and often not addressed how to acquire context in a Ubiquitous
in current research. A key to understanding Computing environment using simple
such systems and their use is the observation sensors is addressed. Then a bottom-up
that humans implicitly interact in context approach for modelling context-aware
with their environments including artefacts is introduced. Prototyping various
context sensing devices and generalizing to
patterns demonstrates the feasibility and communication capabilities. Here the focus
applicability of the approach. Assessment is is to empower humans interacting in such an
further done by evaluating the ways in environment and to enhance their interactive
which providing sensor based context can be experience. This viewpoint is in the tradition
made easier by using methods, libraries, of early interactive computer systems; as an
tools, and physical building blocks. Given early innovator in this area Douglas
that context is made available, the issue of Engelbart, described his work as augmenting
distribution is addressed. From this, a new human intellect contrasting with work that
human computer interaction model that focused on automating tasks [Engelbart,62].
includes context is proposed; a model that Over the last few years there have been
also takes account of explicit and implicit observations of a further shift towards
user interaction. Finally ways are addressed creating a good experience for the user
in which Ubiquitous Computing systems can [Newman,01]. From a technical perspective
be evaluated. The research is motivated by many computer systems have already
the idea of a human centred approach to achieved efficiency.
Ubiquitous Computing, as outlined in the
following paragraphs. The research reported
however includes hardware, software, and
communication issues, as well as topics
related to interaction.

2. Ubiquitous Computing

Approaches in Computer Science in

the last 50 years can be related to the
quantitative relationship between computers
and humans. At the very beginning many
people shared a single computer, then the
idea that each user has a single computer
significantly changed the way people used
computer systems. In the last decade this
changed further into a many-to-one
relationship, where one user has many
computers, or at least devices with
processing capabilities available to, and
surrounding a single user. This recently
started era is referred to as Ubiquitous
Computing; however, Ubiquitous
Computing raises many issues beyond the
quantitative relationship between computer
and user [Weiser,91], [Weiser,96]. From a However, creating a good experience
Human computer interaction (HCI) for a user can improve their perception of
viewpoint Ubiquitous Computing describes their work and so ultimately make the
the phenomenon of interacting in context process of using a computer system more
with artefacts and environments that are effective and pleasurable. This is especially
interwoven with processing and true of tasks where creativity of the user is
essential (e.g. writing, designing, implies that the user is primarily doing a
construction) or where the task itself is task and is not aware of operating a
recreational (e.g. games): In these examples computer system.
experience becomes the main factor. Using
conventional interfaces, such as desktop 3. Interaction in Ubiquitous
computers with screen, keyboard, mouse, Computing
and speakers limits the design space for
creating experience; however, including the
The terms calm computing [Weiser,98],
real environment as part of interaction for a
invisible computing [Norman,98], and
computer system offers many interesting.
disappearing computer [Wejchert,00]
A major challenge in Ubiquitous
describe the user interface perspective on
Computing is physical integration and
Ubiquitous Computing. As the interaction is
embedding of computing and
interwoven with the user’s actions the
communication technology into
concept goes beyond the traditional
environments and artefacts. Such
understanding of a human computer
developments lead to ‘augmented artefacts’,
interface towards describing the relationship
raising issues beyond the physical
between the user and their augmented
integration. Embedding technology into
environment.Making the computer invisible
everyday artefacts also inevitably implies
is not a matter of size or a challenge of
embedding the “computer” into tasks done
seamless integration of hardware, it’s about
by the user. possibilities.
how the human perceives the computer. To
This leads to new research
make the computer disappear (at least in the
challenges and further questions.
user’s perception), the interaction has to be
• What is the consequence of artefacts and
seamlessly integrated with the primary task
environments becoming an integral part
of the user. The user still interacts with the
of the “computer”?
tools that help them to do a certain job, but
• How is it possible and even pleasant to
their focus is on the task itself. This is in
interact with a system where many
contrast to typical usage of a computer as a
artefacts and the environment is a part of
tool, where the focus usually ends on the
the “human computer interface”?
computer not on the task [Weiser,98]. In a
• Where is the application and how do we
Ubiquitous Computing paradigm tools are
influence and interact with an
enhanced with processing and
application when each part of the
communication capabilities to help with
“computer” and of “human computer
achieving the task not drawing focus away
interface” is a potentially a part of many
from it.Embedding interaction into tasks
applications?
seems to be the obvious approach to take.
These issues that are central to research in
However, when it comes to modelling and
Ubiquitous Computing lead to the research
implementing this vision many unresolved
addressed in this thesis. Context, especially
issues appear. Using explicit interaction, as
making context available to the “computer”,
in conversational computer systems, there is
is regarded as a dominant prerequisite to
provision for a choice of varying modalities.
advance on these questions [Abowd,00].
The interaction designer can chose from
In an optimal setting the technology
command line user interfaces, graphical user
disappears so that the “computer” and the
interfaces, and speech and gesture
“human computer interface” are hidden at
interfaces. Independent of the modality the
least in the perception of the human. This
user still is required to interact with a
computer. Another issue that makes reaction that a certain action will provoke is
conversional interaction methods difficult is a major advantage, and at large essential for
that interface components can be physically survival. Many expectations are just
distributed and dependent on each other. On extrapolated from previous experience, e.g.
the other hand as there is also the potential when operating a light switch on a bedside
for many applications to run at the same lamp in a hotel we expect that this particular
time, inputs have to be directed to a light will switch on. We would be rather
particular one. Using solely this approach surprised if instead of the bedside light
would inevitably result in a complex coming on the fan in the shower started or
interface and require a great deal of the the radio in our car outside the hotel starts
user’s attention, which is regarded as one of playing. Our expectations are based on
the most precious resources because it is experience and are essential to the way we
limited. When interaction is embedded it live. In Ubiquitous Computing
happens in context. The physical environments, where the real world becomes
environment, the situation, the role of the a part of the computer and of the user
user, their relation to other users and to the interface, users expectations towards the
environment, and their goals and preferences system are also widely based on the
can all be rich source of information. Using experience of interaction in the real world.
this information when making a system The designer however has a great freedom
context-aware can make the explicit of how to design interaction processes in
interaction process much easier or even such systems. Many limitations inherent in
eliminate the need for explicit interaction. A conventional engineering are no longer an
reduction in explicit interaction will also issue in Ubiquitous Computing, in fact a
reduce the demands on the user’s attention, networked switch could operate anything
assuming that the system gets it right. This else that is networked. To make a system
raises a further issue: how to acquire and useful and give the user the feeling of being
provide context. in charge of the system a switch should
operate what the user anticipates in a
4. Context-Awareness is an particular situation. This simple example of
a switch shows that context is essential for
Enabling Technology
building usable Ubiquitous Computing
systems that respond in a way that is
In Ubiquitous Computing, interaction with anticipated by the user. Context-awareness
computers is inevitably in context and in becomes a fundamental enabling technology
most cases context matters for not only the for Ubiquitous Computing and is a key issue
users directly, but it also matters indirectly when creating computers that are invisible
for the system. The user’s expectations and disappear in terms of the user’s
about a system and their anticipation of the perception. In these terms context-awareness
reaction of a system that they are interacting goes beyond providing context information,
with, is highly dependent on the situation it also requires understanding context and
and environment, as well as on prior ultimately understanding situations.
experience. Interaction in the physical world
is experienced from a very early age and the
5. Challenges in Context-Aware
knowledge about the reaction of the
environment accumulated over a lifetime. Computing
This knowledge allows intelligent behavior;
in particular the ability to predict the
Examples, demonstrators, and prototypes Acquiring context is a prerequisite
have been used to demonstrate that context for any context-aware system. Generally
awareness can enhance applications and context acquisition can be seen as the
systems. Typically location is sensed and process where the real situation in the world
then based on the location further is captured, the significant features are
assumptions about the more general context assessed, and an abstract representation is
are made. As the concept of position and created, which is then provided to
location is well understood, it also provides components in the system for further use.
a powerful and easy to apply model for Approaches to acquire context are manifold
context-aware applications. In many cases and include computer vision, location
however awareness based solely on location tracking, sensor systems, and also more
lacks information that can be of interest to a predictive approaches such as modeling
system for making it context-aware. If users and their behavior.
information beyond location information is
required, further complexity is introduced. • Connecting context acquisition to
context use:
The following issues are central research In a location-aware system there is a close
challenges in context-awareness: relationship between context acquisition and
context use, most often the location sensor is
• Understanding the concept of context: attached to the device using position as
What does context mean and how is context. In this case the context
it connected to situations in the real world? representation is also agreed between these
There is still a fundamental lack of components. In more general environments
understanding in terms how contexts relate context use and context acquisition is
to situations and how general context distributed. It can be assumed that context is
information can be used to help enhance provided for various applications,
applications. This is also associated with the potentially in dynamic configurations. This
question of how to represent context in a makes it obvious that mechanisms to
universal way. connect context acquisition and context use
• How to make use of context? become essential. Here the challenge is
Assuming that context is available in twofold: overcoming the distribution issue
a system the question what is context useful by networking components and agreeing on
for becomes imminent: especially if contexts representations that are useful for a
beyond location and available resources are multitude of components.
considered. In this instance a central
question is what type of applications can be • Understanding the influence on human
enhanced? When considering context as computer interaction:
additional input, issues of reliability and When systems are context-aware their
ambiguity are important. Furthermore the behavior is dependent on the context of use
relation between context and other inputs or the general situation of use. The ultimate
into the system and how they influence each goal is to make systems in such a way that
other, have to be addressed. Ultimately this they react as anticipated by the user. In real
requires the smartness of the system to life however this creates complex problems,
understand the context it is dealing with. in particular if the system reacts differently
from the users expectations. Two critical
• How to acquire context information? issues are how can the user understand the
system and its behaviour? and how to give II. How Ubiquitous computing
the user control over the system? differs from Traditional
computing:
• Support for building context-aware
Ubiquitous computing systems:
These connections are fundamentally
Context-awareness is an enabling
unlike those we associate with networks.
technology for Ubiquitous Computing
Rather than using the network to connect
systems and therefore commonly required
computers that are being used directly by
when realizing such systems. To build
people, these appliances communicate over
Ubiquitous Computing environments
networks such that people do not directly
efficiently, it is inevitable that we need to
monitor the communication between
provide support for building context-aware
machines and programs. The majority of
applications.
these communications will occur in an end-
Up to now, there are many cases where
to-end structure that does not include a
the wheel is re-invented; where all the
human at any point.
problems have to be solved over and over
again in each system. Providing support for
context acquisition, context provision, and
context use will make the process of
implementing context-aware applications
much simpler.

• Evaluation of context-aware system:

As context-aware systems are used in
context, evaluation itself is also required to
be done in context. In cases where
functionality is only available and useful in
a certain context it is required to create or
simulate a particular situation that results in
the wanted context in order to assess the
system. Inducing a particular situation and
context however may have also a significant
effect on measures in the evaluation. The number of machines connected to the
Many of these research issues are Internet has been increasing at an
highly interconnected. Nevertheless some of exponential rate and will continue to grow at
the issues can be tackled fairly this rate as the existing networks of
independently of some others. embedded computers, including those that
In the approach pursued in the course already exist within our automobiles.
of research underlying this thesis, context- The kinds of devices that will be used
awareness is approached from a bottom-up to access the Internet are no longer confined
perspective. In the bottom-up approach to desktops and servers, but include small
context acquisition and context use is related devices with limited user interface facilities
to artefacts. This approach cuts across such as cell phones and PDA’s wireless
several of the challenges above. devices with limited bandwidth, computing
power, electrical power and embedded
processors with severe limitations on the
 amount of memory and computing power all work in this area, but to work
available to them. toward technological, design, and
Many of these devices are mobile, social solutions to address these
changing not only geographic position, but concerns. A basic concern about
also their place in the topology of the any information stored in a
network. computer is knowing who can
access and modify the contents.
Unlike traditional Desktop Computers Where are the bits? Are they
and existing networks, the new devices secure? Security and encryption
will have the following characteristics: schemes are part of the
* Many will have small, inexpensive technological solutions available,
processors with limited memory and little or especially as information is gleaned
no persistent storage. from the environment and
* They will connect to other computing transported over networks.
elements without the direct intervention of Alternatively, work in wearable
users. computing emphasizes a design
* Often, they will be connected by wireless approach—providing security by
networks. keeping the bits local (on the body)
* They will change rapidly, sometimes by and removing the risks of
being mobile, sometimes by going on and transporting them over a public
offline at widely varying rates. Over time, network.
they will be replaced or fail far more rapidly One fear of users is the lack
than is now common. of knowledge of what some
* They will be used as a source of computing system is doing, or that
information, often sending that information something is being done “behind
into the center of the network to which they their backs.” Although the original
are attached. vision of ubiquitous computing
described computing as
III. Social Issues for Ubiquitous disappearing into the physical
Computing environment, this “invisibility” is
counter to informing users about
We are pushing toward making it how they are being sensed. To
easier for computation to sense, assuage that fear, design solutions
understand, and react to can be employed to make this
phenomenon in the physical world information visible.
and to record those phenomena. For example, systems that
These enabling technologies carry sense physical phenomena and
with them numerous dangers, e.g., capture live situations should
making it too easy for people to provide clear indicators that this
build systems that effectively spy sensing or recording is occurring.
on others without any controlling As these sensing and recording
authority. Ubicomp researchers capabilities are more commonly
would be remiss if they undertook found, one challenge for everyday
their work without understanding computing is to enable people to
these issues. However, the fear of be aware of how they are being
wrong-doing is not a call to cease sensed. Just as people can
ascertain their visibility in physical perfect capture system to “forget”
space (How public is this space? some part of the past. An
Are there windows?) we need cues interesting challenge for
to convey our visibility in virtual collaborative situations is to figure
space. out acceptable policies for
The next step is to allow “erasing” or “forgetting” some
those being sensed or recorded to shared memory. A more positive
have control to either stop this slant on this issue would focus on
activity or to at least control the ways to accommodate heightening
distribution and use of the awareness of particularly valuable
information. This challenge is segments of a captured experience
related to the design of in lieu of eliminating or forgetting
collaborative environments where parts of a captured history.
the actions and roles of Although issues surrounding
collaborators are fluid and difficult the appropriate use and
to articulate in a static snapshot. dissemination of information are as
The capture, distribution, and use old as the dawn of human
of information will be determined communication, specific concerns
over time by the specific practices stem from ubicomp making a new
of people is workplace and home kind of information more generally
settings. available.
There are a number of On the positive side,
reactions that system builders can knowledge of recording encourages
have for handling the sensitive people to be less reckless in their
topic of when and what to capture. commentary. On the negative side,
At Xerox PARC, one solution for this same knowledge can cause
capture was to agree to only people to refuse to contribute to a
capture the summary portions of discussion for fear of saying
technical meetings. In Classroom something that would be regretted
2000, we defaulted to recording all in the future.
of a lecture, but did not attempt to A more subtle problem was
obtain good quality audio or video noticed in our extensive experience
of anyone except the lecturer in in Classroom 2000. Some students
the front of the room. In the indicated that they chose not to
Dynomite system from FX-PAL, the ask questions in class because the
note-taker exercised control over answer was likely already
which portions of the audio discussed and it was up to the
recording could be kept for future student to go back and listen to the
reference [Wilcox et al. 1997]. lecture.
Though this last solution was In general, social and legal
first presented as a means of practices continue to evolve in
reducing the amount of storage concert with technological and
requirements for high-fidelity design innovations. In each
audio, we see merit in this situation people will compare the
approach from an individual’s perceived benefits and costs of the
perspective to enable an otherwise uses of ubicomp technologies. For
 example, skiers and hikers choose constantly encountered. Ubiquitous
to wear radio transponders so they computing, or “ubicomp”, as it has been
can be located by rescue tagged will define the future of computing.
personnel. Firefighters benefit from The distinguishing feature of
understanding what each of them
ubicomp will be the lack of interface.
is doing and where they are
located. Recent research details Everything will be controlled by natural
the calendaring practices at Sun actions as opposed to the point-and-click
Microsystems [Grudin and Palen interfaces we have all grown used to. Right
1997] where colleagues share now we receive information in two distinct
extensive information about their ways: pull or push. “Pull” can be
daily collaborative activities. characterized by a user sitting down at a
As discussed in the previous
computer, firing up Google, and searching
section on evaluation, our
understanding of the social for specific information in real-time. “Push”
implications of these technologies is characterized by receiving filtered
will often come after people invent information based on user preferences;
new, unforeseen, uses of these much like the personalized text messages on
technologies. Although the sand is your cell phone informing you of weather or
always shifting beneath us, traffic conditions. When computing
attention to issues of security, becomes ubiquitous you will not need to
visibility, control, and privacy
manually set preferences.The object you
should help ensure a more positive
use of these technologies. interact with will learn from you and
provide information based on your
The Future: Ubiquitous Computing environment. Temperature, time of day,
movement, sound, color and light will all
Computing, over the past 50+ years, has influence the information you receive.
gone through two distinct phases: the Ubiquitous computing will provide a
mainframe era and the personal computing continuous stream of information without
era. The third phase has begun and you may being distracting and will only provide the
have not even noticed – that is the way it is information you need at the time.
supposed to be. Ubiquitous means existing Everything will become interactive and
or being everywhere at the same time, i.e. more importantly, reactive.
Imagine the following scenarios: computing will probably prevent most
accidents before they happen.)
1. You make a call to your friend whose 3. You have lost your keys. No more
native language is French. He searching, just ask your house. It will
understands English quite well but prefers know EXACTLY where they are, even if
to speak in French. No problem. In real- they are hiding in the couch cushions.
time what you say comes across on his (Keys will probably be a thing of the past
end in French and vice-versa. at this point.)
2. You are rushed to the hospital after a car
accident. By performing a retinal scan the It is difficult to comprehend all that
ER doctors are provided with time- ubiquitous computing will entail. I look at it
sensitive and important information: this way: Everything, and I mean
allergies, past surgeries, existing everything, will be connected. A
conditions, emergency contact
communication device of some type will be
information, name and age. (Ubiquitous
embedded in every single product created.
 This prospect is scary for some, exciting for MyGrocer Ubiquitous Computing
others. Environment First International Mobile
Business Conference, 150-172.
[8]. Weiser, M. (1994). The world is not a
IV. Applications of Ubiquitous
desktop. Interactions, 1 (1), 7-8.
computing: [9]. Weiser, M.(1998) The invisible
interface: Increasing the power of the
They are applicable in cars, mobile phones, environment through calm technology.
handhelds, households appliances, smart CoBuild, 1.
machines, wearable computers and so on... [10]. Nigel Davies, Hans-Werner Gellersen
"Beyond Prototypes: Challenges in
V. Conclusion: Deploying Ubiquitous Systems" IEEE
Pervasive Computing, January-March
We are moving toward a future in which 2002 (Vol. 1, No. 1) pp.26-35
computing will be ubiquitous, woven
seamlessly into the fabric of everyday life.
Researchers are engaged in several projects
to explore technologies and usage models
for everyday uses of computing. In their
research, they are addressing fundamental
issues that must be resolved.

VI. References

[1]. Abowd, D., Mynatt, E.: Charting Past,

Present, and Future Research in
Ubiquitous Computing. ACM TOCHI,
Vol. 7, No 1 (2000), 29-58.
[2]. www.pervasive.ifi.lmu.de/
[3]. Shilt, B. N., Adams, Norman A., Want,
R. (1994). Context-aware computing
applications. IEEE Workshop on
Mobile Computing Systems and
Applications.
[4]. Coatta, T. (2006). The (not so) hidden
computer. ACM Queue, 4 (3), 23-26.
[5]. Weiser M (1991) The computer for the
21st Century. Sci Am
[6]. 265(3):94–104Weiser, M. (1991). The
computer for the 21st century. Scientific
American, September, 1991, pp. 94-
104.
[7]. Kourouthanasis, P., Spinellis, D.,
Roussos, G. & Giaglis, G. (2002).
Intelligent Cokes and Diapers: