The guiding concepts and models used to drive the design and implementation of ubiquitous computing and its technology infrastructure and associated applications will largely determine its character. RFIDs, smart sensors, back-end servers and wireless access to information everywhere are only the technology components of a system. How these are integrated with each other, what form they take, what and where the human interface is, and who controls their capabilities, behavior and information content is still to be determined.
The Internet is an instructive case study for how a good model benefited an emerging medium. The strong vision and guiding concepts defined by ARPA (later renamed DARPA – Defense Advanced Research Projects Agency) with its contractors the Rand Corporation and BBN Technologies formed an open platform that has accommodated many productive uses that were completely unanticipated when the Internet was created.
While driven by Cold War concerns of survivable operations in the case of a nuclear attack, the radical model (especially for a centralized command-and-control military) of a decentralized network where information is broken down and passed around using open standards has turned out to be incredibly robust and useful. The model created the opportunity for applications that include email, the Web, Internet telephony and self-published video, all running on a system design whose basic philosophy was established in the 1960s. The productiveness of the system grows directly out of the model defined and defended by the Internet’s founders.
So now, in the early days of the 21st century, we are at a moment when designers, engineers and, I hope, the public, have an opportunity and the responsibility to define beneficial models for an emerging new medium of technologies integrated into everyday things. Will it have the open and productive character of the Internet? Will it have the closed, centrally controlled model of cable TV? Or are there other models for this new medium? For an understanding of possible directions, we can look at some of the visions given to this new infrastructure. Proposed models include those designated by a variety of terms, including ubiquitous computing, ambient intelligence, everyware, sensor networks, the Internet of Things, and our term, the New Ecology of Things.> 2. Ubiquitous Computing
By entering a room, you trigger a cascade of responses on the part of embedded systems around you. Sensors in the flooring register your presence, your needs are inferred (from the time of day, the presence of others, or even the state of your body), and conditions in the room alter accordingly.
Speech, too, carries clear cues as to the speaker’s emotional state; a household system might react to these alongside whatever content is expressed – yes, the volume can be turned down in response to your command, but should the timbre of your voice indicate that stress and not loudness is the real issue, maybe the ambient lighting is softened as well.
In this view, embedded systems use complex rules to behave without intervention. The systems sense characteristics of people and the environment, infer needs and make automated decisions that cause changes in the world, turning on lights, collecting and presenting information, or purchasing products and services.
While the automatic accommodation of all our needs seems seductive, there are some questionable assumptions in this approach. First is the idea that systems can always be smart and helpful enough that people won’t need to participate in the decision-making process. This idea has been around for a long time, in visions of the house of the future, for example, where refrigerators automatically buy groceries as needed, and, as in Greenfield’s scenarios, environmental controls that adjust to our moods and make us happier. This position has a kind of utopian optimism similar to the early days of the Web, when people believed, for example, that brick and mortar retail stores would be eliminated by online shopping, and loneliness would be dissolved through virtual communities.
The disappointing history of artificial intelligence tells us that ubicomp systems may never be very sophisticated in their ability to understand the intentions, meanings, emotions and desires inherent in the ambiguous, everyday actions and speech of people in homes, public spaces and work settings. But more importantly, do people really want this kind of automation? For example, do you want a system deciding what should happen when you are stressed? One day, you may want to hear the band Zero 7 to relax. Another day, you may decide you want to listen to Metallica. Or one week you may want to buy strawberries because they are locally in season, and the next you may want imported peaches for a special occasion. In human terms, will it be very satisfying or useful to have your stereo or refrigerator automatically make your music or food choices?
People often decide what they want by exploring possibilities, learning as they go. We like options, and often make new discoveries by trying out different things and taking advantage of serendipity. The model of ubiquitous computing assumes that the “right” outcomes can be determined in advance, and decisions logically flow from assumptions. But in the real world we want more flexibility. The invisible, ambient character of ubicomp does not provide the affordances or openness that encourage or even accommodate a lively sense of discovery and productive, meaningful creation. Where are the choices and the interfaces for them?
This idea of “right” outcomes raises further questions. Who controls how these systems make their decisions? How do we influence the algorithms and assumptions? Will our “computationally enhanced” lives be dominated by organizations like our cable companies who limit our lives to four or five demographically optimized packages? It’s one thing to have TV channels packaged, but imagine living with a pervasive home-automation system that’s programmed and managed by Comcast or Clear Channel. With a model such as ubiquitous computing where the system exists under the surface of life, what openness and user agency is lost? Do our lives become centrally controlled, over-systematized and predictable? What values are emphasized or suppressed?> 3. The Internet of Things