What is Ubiquitous Computing?
Ubiquitous computing is the act of using computational power without being explicitly aware of the computers themselves. Mark Weiser's work at Xerox PARC pushed, "a new way of thinking about computers, one that takes into account the human world and allows the computers themselves to vanish into the background." (Weiser, 1991) As people learn things sufficiently well, they cease to be aware of them. To a large extent, current technology leaves people chained to their desktop computers. As long as home computers are positioned as a form of high-tech shrine, they can hardly be considered personal. There are two main driving forces which will cause computers to become ubiquitous as time passes. First, we will learn to ignore their presence, just as we have accepted telephones; simultaneously, their presence will be minimized as they themselves physically shrink to the point of concealment.
Engineering developments in electronics, materials, and many other fields are providing the fuel for cancelable computers - where the "footprint" is small or flexible enough that even direct attention to a given artifact does not necessarily reveal that it is a computer. In the interest of accepting computers into our daily activities, numerous scientists and developers strive to, like Dr. G. Abowd, "provide an interface that can take on the responsibility of locating and serving the user." (Abowd et al., 1997) At this point, a logical yet somewhat artificial classification system is worth introducing because the sensors and services provided by a system depend to a large extent on the number and "proximity" of the users.
Wearable vs. Fixed Computing Systems
Different sizes of computers can be equipped with various sensing devices. In general, sensor data concerning a large space (like a room) is most easily gathered (and represented) by stationary systems built into the surroundings they are monitoring/presenting to. Similarly, data can be gathered about and relayed to a person most conveniently when that person is wearing the necessary computer system. As ubiquitous computers are developed some will be wearable, some will not, and some will blur the distinction altogether. Fundamentally though, the feedback is directed to either an individual - often stimulating an augmented reality, or to a group of people - customizing their computational services.
I. Ubiquity you wear: Augmented Reality
Augmented reality is a relatively new term. It can apply to high-tech gadgets or to "inventions" more than 100 years old. I take it as referring to our ability to continuously supplement our natural (naked?) knowledge with extra information, which we would otherwise lack. Arguably, eyeglasses are such a device since they allow us to see differently. More modern is the example of the wrist- or pocket watch. Without these, out perception of time would be based on the sun or moon, and likely inaccurate.
There is a strong relationship between augmented reality and wearables because our perception of reality must be continuous. Reality augmented with the help of computers generally requires some form of wearable computer. A wall clock ceases to change our perception of time when we walk away from it, but a watch worn with the body is always available. This may seem like a subtle distinction, but limited wearability can cause us to leave even extremely useful devices out of immediate reach. Binoculars, like glasses, enhance our ability to see, yet far fewer people carry such bulky objects along regularly.
Wearable computers in 1997 are surely much more bulky than they will be in the coming years. As much as people wear different types of clothing, any two wearable computers are likely different, depending on their uses as well as their wearers’ tastes. In general, a "wearable" tends to consist of some sub notebook sized central computer which drives a translucent computer display or some form of monocular display. The computer is controlled by peripherals such as joysticks, chording-keyboards, 3D mice, or by nontraditional means such as gesture or speech recognition. A wearable becomes significantly more powerful (and maybe intrusive…) when it has sensors that help it detect the user’s location, situation, mood (Picard & Healey, 1997), and even intentionality. Typical sensors include digital cameras, GPS units, thermometers, and microphones. Until wearables shrink further in size and conspicuousness, persons making use of their augmented reality look as weighted down as their counterparts experiencing virtual reality.
Virtual reality should not be confused with augmented reality. Both
may require artifacts, but virtual reality (also a broad topic) replaces
our natural surroundings with different and usually computer-generated
ones. Augmented reality never tries to trick a person into thinking that
they are somewhere else. Virtual reality causes a person to see the world
through a computer's perception, fooling the natural senses. In contrast,
augmented reality only depends on computers to supplement our natural senses,
attempting to enhance them.
Ia. Why wear it?
Arguably, humans are fantastic "machines," adapting and evolving to
match their needs and their surroundings. Technology often attempts to
emulate our human skills of motion, perception, and reasoning. These attempts
generally yield results that are far inferior to our natural skills, so
one could ask what technology could possibly offer in terms of enhancements.
Wearables can be grouped as addressing two main needs; overcoming handicaps
and personal information storage/management.
As wonderful as our bodies are, parts are sometimes damaged or inoperative. The development of prosthetics serves to restore some of our "normal" motor functionality. Some handicaps require more advanced modifications, warranting the involvement of technology, potentially in the form of a wearable computer system.
The classic case for wearable computers is made in the treatment of visually impaired persons. Advocates of wearables at MIT explain that, "Approximately 2 million Americans are affected by low vision, a set of conditions which can not be corrected with normal eyeglasses and severely affects the individual's sight. However, some of their needs may be addressed by remapping the visual input." (Starner et al., 1997) For example, by connecting a camera and a display system to a wearable computer, it is possible to digitally magnify a section of a person’s visual field, allowing them to see somewhat normally. Similarly effective is the remapping of the visual field around scotomas (blind spots).
Just as human limitations in vision can be accommodated, so too can our audio perception. Hearing aids have become quite accepted as tools for presenting an augmented version of our surroundings. With new developments of mini-speakers, persons with neurologically limited short term memory can have parts of a recent conversation repeated back to them, without anyone else even noticing such a "hearing aid." Also. speech synthesis software is being combined with gesture recognition algorithms to build a baseball cap mounted translator for sentence level American Sign Language (ASL) (Starner, Weaver, et al., 1997). In this case, a wearable system is augmenting the space around the user, as well as the user’s own reality. Consequently, a person who would normally only be understood by an audience of ASL "speakers," can be understood by listeners who speak any spoken language the speech system is designed to generate.
Spinal column injuries are among the most debilitating, often resulting
in permanent paralysis of arms and/or legs. As impressive as modern medicine
has become, patients with damaged spines still face the fundamental challenge
of controlling their environment, including what are otherwise perfectly
functional limbs. A wearable speech recognition system is one possible
instrument for restoring the person’s control. Research work being performed
by Dr. Phil Kennedy (Kennedy,
1997) in the area of monitoring the central nervous system has the
potential of evolving into a wearable actuation system. Electrodes permanently
attached to the brain are being used to generate radio signals which are
received by a hat. This research suggests that eventually, a computer could
translate those signals into meaningful gesture and motion commands - at
least partially restoring the person’s control of their surroundings.
Ic. Information Storage & Management
Human memory is a volatile system. We have trouble committing certain facts, numbers, and events to memory. For this reason, we keep notes in calendars, address books, on our computers, and lately - on Personal Digital Assistants (PDAs). This behavior does not indicate a human infirmity - simply a preference for committing to memory more valuable and complex information. Once date is stored, it is nontrivial to retrieve it when it is appropriate and meaningful. PDAs used as portable computers are relatively simple devices from which a user must extract information by knowing where to look for it (under what headings). By incorporating this portable computing into a wearable system, we convert our memory aid into an active portion of our reality. The MIT Media Lab has been developing a wearable "Remembrance Agent" which serves as a system for augmented memory (Rhodes, 1997). It is active because it uses what contextual information it can obtain from its sensors to volunteer information which could be useful in a given situation. When the sensors correctly identify a situation, the user sees overlaid text or hears ambient "earcons" which relay the relevant information. These systems are often modular and lead to personalized types of augmented memory. Some people may use this system purely to record and retrieve address and calendar data, while others recall face-name associations and prefer to see maps overlaid on their field of view when they are in a new city (Abowd et al., 1997).
Beyond what users may have already stored with their remembrance agents, they may need more information. Again relying on the wearable system’s sensors, the wearable can display "physically based hypertext." (Starner et al., 1997) These overlays serve as links to further information which may be stored locally (in the wearable computer) or may be retrieved wirelessly from a database or network. Airline mechanics would have difficulty consulting the 2 ton manual associated with 747’s, but can access any portion of it quickly using a wearable which picks up enough cues to recognize which part of the airplane is being inspected.
Id. Links
II. Ubiquity you Don’t Wear: Tabs, Pads, and Boards:
Instead of augmenting a single person’s reality, a ubiquitous computer
design can also be used on multiple persons at once. Weiser and his group
at Xerox PARC built what they
called tabs, pads, and boards. These artifacts also have low-tech equivalencies:
tabs
are like Post-it notes, Pads are like legal-sized paper pads or magazines,
and boards
are essentially extensions of the white/blackboards we know from school.
 |
 |
They are "smarter." Credit cards, access cards, and identification badges are already accepted as normal artifacts. These are in fact, inferior to tabs, because they have no association with their context: A stolen credit card will continue to operate, as will the other cards. We already have the hardware capabilities of smartcards, so the concept of tabs simply pushes for intelligent use of those capabilities - like code- or fingerprint-confirmation. Anytime people need to communicate in a way other than purely verbal, they tend to use media tools. From cave-wall etches to CAD drawings, from simple grunts to complex speech, information quality is closely tied to the form in which it is presented. Paper is expensive and it can be damaged or lost. Tabs, pads, and boards represent the merging of computer technology with old-fashioned paper. Among the differences: now information is easily stored and disseminated, it can be easily searched and retrieved. Computer companies such as IBM have made a business of providing central mainframes to companies which feel that this is a good way to store their information. Now we are examining whether adding ubiquity to this formula improves our communications.
IIb. Non-wearables in Education
Different people learn differently - some are visual learners, some rely on what they hear, and some just learn by doing. To accomodate as many learning styles as possible, researchers in the Future Computing Environments group at Georgia Tech are developing a project called Classroom 2000. The project focuses on techniques to capture, index, browse, and retrieve various digital media which is relevant to given course material. Students find different aspects of the available data more or less useful, depending on their personal learning styles. I made use of Dr. Gregory Abowd's lecture notes on Ubicomp in preparing this document.
IIc. Non-wearables for All Occasions
The Future Computing Environments group at Georgia Tech has numerous projects in various stages of completion which illustrate some of the new areas ubiquitous computing will explore. As Mark Weiser himself points out - predicting ubiquity is by nature an impossible task.
The day has not yet arrived where everyone enjoys the benefits of computers
without being troubled and aware of their presence. However, wearables
are already mobile and practical enough that certain people can benefit
from their use, either for physical reasons, or as a means for increased
efficiency. Also, non-wearables are finding great acceptance either because
we are getting used to them, or because we can't actually see them, or
both. The future of ubiquitous computing will likely affect various aspects
of daily life in a manner parallel to the telephone; It is both wearable
and non-wearable, but ubiquitous nonetheless. It's "future" lies in it's
context-sensitivity to its surroundings, resulting in personal calls being
routed to that person or the space they occupy (whichever is better). When
this "behavior" becomes natural to us, we will have yet another example
of ubiquitous technology.
Abowd, G., Dey, A., Orr, R., Brotherton, J., (1997). Context-awareness in Wearable and Ubiquitious Computing. In Proc. Of International Symposium on Wearable Computers, pages 179-180, Cambridge, Massachusetts.
Kennedy, P. (1997). The Development of Direct Interfaces with the Central Nervous System for Spinal Cord Patients. http://www.cc.gatech.edu/gvu/info/brownbag/971009.html.
Picard, R., Healey, J., (1997). Affective Wearables. In Proc. Of International Symposium on Wearable Computers, pages 130-137, Cambridge, Massachusetts.
Rhodes, B. (1997). The Wearable Remembrance Agent: A System for Augmented Memory. In Proc. Of International Symposium on Wearable Computers, pages 123-128, Cambridge, Massachusetts.
Ross, D., Sanford, J. (1997). Wearable Computer as a Remote Interface for People with Disabilities. In Proc. Of International Symposium on Wearable Computers, pages 161-162, Cambridge, Massachusetts.
Starner, T., Mann, S., Rhodes, B., and Levine, J. (1997). Augmented Reality Through Wearable Computing. Technical Report 397, MIT Media Lab, Perceptual Computing Group.
Starner, T., Weaver, J., and Pentland, Alex, (1997). A Wearable Computer Based American Sign Language Recognizer. In Proc. Of International Symposium on Wearable Computers, pages 130-137, Cambridge, Massachusetts.
Weiser M., (1991). The
Computer for the 21st Century. Scientific American, 265(3):94-104.