Thesis 03


Thesis 03

Everyware is information processing embedded in the objects and surfaces of everyday life.

Part of what the everyware paradigm implies is that most of the functionality we now associate with these boxes on our desks, these slabs that warm our laps, will be dispersed into both the built environment and the wide variety of everyday objects we typically use there.

Many such objects are already invested with processing powermost contemporary cameras, watches, and phones, to name the most obvious examples, contain microcontrollers. But we understand these things to be technical, and if they have so far rarely participated in the larger conversation of the "Internet of things," we wouldn't necessarily be surprised to see them do so.

Nor are we concerned, for the moment, with the many embedded microprocessors we encounter elsewhere in our lives, generally without being aware of them. They pump the brakes in our cars, cycle the compressors in our refrigerators, or adjust the water temperature in our washing machines, yet never interact with the wider universe. They can't be queried or addressed by remote systems, let alone interact directly with a human user.

It's not until they do share the stream of information passing through them with other applications and services that they'll become of interest to us. It is my sense that the majority of embedded systems will eventually link up with broader networks, but for now they play a relatively small role in our story.

By contrast, what we're contemplating here is the extension of information-sensing, -processing, and -networking capabilities to entire classes of things we've never before thought of as "technology." At least, we haven't thought of them that way in a long, long time: I'm talking about artifacts such as clothing, furniture, walls and doorways.

Their transparency is precisely why this class of objects is so appealing to engineers and designers as platforms for computation. These things are already everywhere, hiding in plain sight; nobody bats an eyelash at them. If nothing else, they offer a convenient place to stash the componentry of a computing power that might otherwise read as oppressive. More ambitiously, as we'll see, some designers are exploring how the possibilities inherent in an everyday object can be thoroughly transformed by the application of information technologies like RFID, Global Positioning System (GPS), and mesh networking.

The idea of incorporating digital "intelligence" into objects with an everyday form factorindustrial-designer jargon for an object's physical shape and sizeappeared early in the developmental history of ubicomp. As far back as 1989, Olivetti Research deployed an early version of Roy Want's Active Badge, in which the familiar workplace identity tag became a platform for functionality.

Workers wearing Active Badges in an instrumented building could automatically unlock areas to which they had been granted access, have phone calls routed to them wherever they were, and create running diaries of the meetings they attended. They could also be tracked as they moved around the building; at one point, Olivetti's public Web site even allowed visitors to query the location of an employee wearing an Active Badge. And while the intent wasn't to spy on such workers, it was readily apparent how the system could be abused, especially when the device responsible was so humble and so easy to forget about. Original sin came early to ubicomp.

Want went on to join Mark Weiser's team at PARC (Palo Alto Research Center), where he contributed to foundational work on a range of networked devices called "tabs," "pads," and "boards." As with Active Badge, these were digital tools for freely roaming knowledge workers, built on a vocabulary of form universally familiar to anyone who's ever worked in an office: name tags, pads of paper, and erasable whiteboards, respectively. [*]

[*] These form factors had been looming in the mass unconscious for a long time. PARC's "pad," in particular, seemed to owe a lot to the slablike media/communication devices used by astronauts Frank Poole and Dave Bowman in Stanley Kubrick's 1968 classic 2001: A Space Odyssey.

Each had a recognizable domain of function. Tabs, being the smallest, were also the most personal; they stayed close to the body, where they might mediate individual information such as identity, location and availability. Pads were supposed to be an individual's primary work surface, pen-based devices for documents and other personal media. And boards were wall-sized displays through which personal work could be shared, in a flow of discovery, annotation and commentary.

Networking infrastructure throughout the office itself enabled communication among the constellation of tabs, pads and boards in active use, allocating shared resources like printers, routing incoming e-mails and phone calls, and providing background maintenance and security functions. Documents in progress would follow a worker into and out of meetings, up onto public boards for discussion, and back down to one's own pad for further revision.

Part of the reasoning behind this was to replace the insular, socially alienating PC with something that afforded the same productivity. In this, PARC anticipated by half a decade the casual, and casually technical, workspace that did in fact emerge during the late-1990s ascendancy of the dot-coms. At least in theory, by getting people out from behind their screens, tabs and pads and boards lent themselves to an open, fluid, and collaborative work style.

Although none of these devices was ever commercialized, at least by Xerox, the die had been cast. Many of the ubicomp projects that followed took PARC's assumptions more or less as givens, as researchers turned their efforts toward enabling the vision of collaborative, distributed work embedded in it.

But what about that percentage of our lives we spend outside the confines of work? While it was more or less inevitable that efforts would be made to provision objects outside the workplace with a similar capacity for digital mediationif for no other reason than the attractively high margins and immense volume of the consumer-electronics sectorit took longer for them to appear.

To understand why such efforts took so long to get off the ground, it's necessary to reconstruct for a moment what the world looked like at the very dawn of ubicomp. As strange as it now seems, the early conceptual work in the field happened in a world without a Web or, for that matter, widespread adoption of mobile phones in North America.

The 802.11b standard we know as Wi-Fi, of course, didn't yet exist. you couldn't simply cobble together a project around off-the-shelf wireless routers. The in-building wireless network prototypes like Active Badge depended on were bespoke, one-off affairs; in more than one project, students simply sketched in connectivity as a black box, an assertion that if an actual network were somehow to come into existence, then the proposed system would function like so.

In such an environment, it may have been reasonable to posit a pervasive wireless network in the workplace. However, a deployment in public space or the domestic sphere was clearly out of the question.

The mass uptake of the Internet changed everything. What would have seemed fanciful from the perspective of 1992 became far more credible in its wake. As a lingua franca, as an enabling technology, and especially as an available kit of parts, the dramatic, global spread of Internet Protocol-based networking immediately made schemes of ubiquity feasible.

Over the next several years, a profusion of projects explored various strategies for living with, and not merely using, information technology. Some of the proposals and products we'll be encountering in this book include keys and wallets that locate themselves when misplaced; a beer mat that summons the bartender when an empty mug is placed upon it; and a bathtub that sounds a tone in another room when the desired water temperature has been reached. [*]

[*] Could the mental models attached to such familiar forms unduly limit what people think of to do with them? The answer is almost certainly yes; we'll take up that question a bit later on.

Some of the most beautiful everyware I've seen was designed by former PARC researcher Ranjit Makkuni, whose New Delhi-based Sacred World Foundation works to bridge the gap between technological and traditional cultures. This is information processing interwoven with the familiar daily forms not of the developed world, but of the global South, cycle rickshaws, clay pots, and amulets among them. It's a lovely reminder that the world contains a great many different "everydays," beyond the ones we happen to be used to.

Whether clay pot or beer mat, though, these projects all capitalize on the idea that the distinctly local application of intelligence, and not the generic, one-size-fits-all vision embodied in computers, will turn out to be among the most important and useful legacies of our technological moment. In this, they appear to be following the advice of human interface pioneer Don Norman.

Norman argues, in The Invisible Computer and elsewhere, that the difficulty and frustration we experience in using the computer are primarily artifacts of its general-purpose nature. He proposes that a truly human-centered design would explode the computer's many functions into a "quiet, invisible, unobtrusive" array of networked objects scattered throughout the home: simple, single-purpose "information appliances" in the form of shoes, bookshelves, even teddy bears.

Or we could go still deeper "into the woodwork." Stefano Marzano points out, in his introduction to Philips Electronics' 2000 exploration of wearable electronics, New Nomads, that when we remove the most transient layer of things from the environments we spend our lives in, we're left with nothing but the spaces themselves, abstracted down to their essentials. These are universals humans have lived in for millennia, elements like walls and roofs, tables and seating, clothing. And, of course, the body itselfour original and our final home. In everyware, all of these present appealing platforms for networked computation.

Fifteen years downstream from its tentative beginnings at Olivetti, the idea of the ordinary as a new frontier for computing is finally starting to bear fruit. We're beginning to see the walls and books, sweaters, and tabletops around us reconsidered as sensors, interface objects, or active sites that respond in some way to data they receive from outside. Eventually, we may even come to see them as the articulated parts of a massively distributed computational engine.

When everyday things are endowed with the ability to sense their environment, store metadata reflecting their own provenance, location, status, and use history, and share that information with other such objects, this cannot help but redefine our relationship with such things. We'll find our daily experience of the world altered in innumerable ways, some obvious and some harder to discern. And among the more significant consequences of this "computing everywhere" is that it strongly implies "information everywhere."