A Chip in Every Pot. Engineers Can Put a Computer in Every Home Appliance, but Why?

By LISA GUERNSEY
The New York times, April 13, 2000

                    Russell Robertson was grappling with an unusual
                    assignment.

              As an industrial designer, his mission was to figure out
              how kitchen appliances will be designed when, as he put
              it, "the fridge talks to the coffee pot."

              His eyes twinkled as he spoke, but he was not kidding
              about the basic concept. In fact, while the idea may have
              once sounded ridiculous, predictions of the advent of
              such devices are now becoming almost clichéd.

              News articles champion the coming era of "smart"
              appliances and "pervasive" or "ubiquitous" computing.
              Engineers have demonstrated how once-dumb machines,
              if embedded with tiny computers, might communicate
              wirelessly and embark on tasks like making coffee or
              ordering groceries. Some manufacturers, like Merloni, an
              Italian company, have already started to sell a few
              products. It appears that the age of "things that think," a
              phrase that emerged from the Massachusetts Institute of
              Technology's Media Laboratory in 1995, is nearly upon
              us.

              Or is it? At ECCO Design in Manhattan, where Mr.
              Robertson works, that was the question behind four
              months of research commissioned by an appliance
              manufacturer. Sketches on the walls behind him showed refrigerators and microwave ovens,
              some shaded with colored pencil and surrounded by scribbled notes. Binders stuffed with
              survey results sat on his desk.

              The project, said Eric Chan, ECCO's founder, had caused his company to ask some
              fundamental questions: Do people need all this automatic information? Do they want
              technology to be that pervasive?

              Others are starting to wonder the same things, whether they are engaged in idle conversation
              or seriously pondering the questions in company meetings. Manufacturing companies are
              hiring design firms and consultants to research whether people would use something like a
              networked washing machine or a refrigerator that knows when they're out of milk.
              Anthropologists who understand consumer behavior are in sudden demand.

              Skeptics, too, are starting to be heard above the trumpeting of pervasive-technology's virtues.
              At a recent Intel conference on product design, some attendees belittled the concept as nothing
              more than clever marketing by companies trying to sell new products. In NetFuture, an online
              newsletter about technology and social responsibility, Langdon Winner, a political science
              professor at the Rensselaer Polytechnic Institute, described most pervasive-computing devices
              as "sheer silliness" and criticized researchers for spending more time trying to make
              appliances talk than attempting to mend society's problems.

              "A lot of ubiquitous-computing development is a tool in search of a purpose," Mr. Winner
              said in an interview.

              Of course, new technology has always drawn skeptics -- even the wheel probably had its
              doubters.

              And questions about whether people really need some sort of new technology can seem
              pointless in a historical context. Who imagined 30 years ago that computers would seem so
              essential or that microwave ovens would be in practically every kitchen?

              And how many people would have forecast that the videophone, a technology that had been
              heralded by futurists for decades, would turn out to be such a dud among consumers?

              But predicting whether a technology will be adopted is critical for companies that want to
              succeed, or even survive, in the marketplace. The ones that can figure out what will be deemed
              useful, superfluous or downright ridiculous will win. And today, as tiny, wireless computer
              systems are being perfected and the Internet is allowing the distribution of data in seconds,
              dozens of appliance manufacturers are betting that some sort of pervasive-computing devices
              will come to be considered as necessary as a telephone. The trick, for them, is to figure out
              which ones.

              A conference earlier this year in Gaithersburg, Md., gave some indication of how much
              engineers and computer scientists are convinced that networked communications devices will
              be mainstays of the future. The conference, called Pervasive Computing 2000, was sponsored
              by the National Institute of Standards and Technology and considered the development of
              standards for such devices. On two snowy days in January, nearly 200 engineers crowded
              into a dimly lit conference room to discuss topics like sensors that could recognize
              conference attendees and home appliances that could alert manufacturers when a repair was
              needed.

                                                        One of the speakers was William
                                                        Mark, vice president for information
                                                        and computing sciences at SRI
                                                        International, which specializes in
                                                        developing new technologies. He talked
                                                        about the complications that could arise
                                                        once appliances were asked to
                                                        communicate. Each machine has its
                                                        own language, he said, displaying
                                                        slides with cartoon images of a toaster
                                                        and refrigerator. A toaster would know
                                                        only "toasterese": signals like "heat" or
                                                        "brown." Similarly, a refrigerator
                                                        would know only "fridgese."

                                                        Somehow, he said, the two would have
                                                        to use an intermediary language to
                                                        communicate. And they might also
                                                        need a language to communicate with
                                                        humans.

              "In the future," Mr. Mark said, "it's going to be quaint to think of a time when refrigerators
              didn't talk to you." He ended his talk with a slide showing a simple equation: "Billions of
              people plus billions of devices equals a community, not a rabble."

              In the audience, Jeffrey Senn, chief technologist for the Maya Design Group, a software
              company in Pittsburgh, shook his head. When asked why he seemed to disagree with the
              presentation, Mr. Senn said: "Not billions of devices. Trillions of them."

              Throughout the presentation, and throughout the conference, the question of why people will
              need toasters that talk to refrigerators was never broached. And to Mr. Senn, such questions
              are not really pertinent. "Creating the raw technology" is the engineers' objective, he said.

              The market will answer the question of whether it is needed or not, he said.

              That is the kind of response that Stephen Talbott, the editor of the NetFuture newsletter,
              considers a "disastrous abdication of moral responsibility." Scientists who refuse to look at
              the consequences of their inventions or who resist questions of societal need are engaging in
              a "willful blindness that cannot lead to good for society," he said.

              Questions about what is best for society, he argued, should come before questions of
              technological standards or the adoption of new technology. Even asking whether something
              would be useful might not be enough, he said. "We should be less concerned about
              embedding little bits of silicon in everything and more concerned about embedding ourselves
              more deeply in real world contexts," he said.

              Dr. Winner, at Rensselaer Polytechnic, said he wished that engineers would consider the
              unintended effects of new technologies before creating devices that seemed to promise to save
              time or reduce effort.

              "As people add more and more time-saving, labor-saving equipment to their homes," Dr.
              Winner wrote in an issue of NetFuture, "their lives do not become simpler and easier. Instead
              their days become even more complicated, demanding and rushed." The latest push to embed
              chips and communication devices into appliances, Dr. Winner said in an interview, is little
              more than serving "the exotic needs of the world's wealthiest people."

              At the M.I.T. Media Lab, where many of today's ideas about pervasive computing were born,
              such criticisms sting.

              Neil Gershenfeld, a professor in the laboratory and the author of "When Things Start to
              Think" (Henry Holt & Company, 1999), wonders if people who ask about the need for
              embedded computers get the point. He envisions pervasive computing, he said, as a way to
              make people's lives better, not worse.

                                                 A few years ago, for example, he developed a
                                                 smart chair that sensed even the smallest
                                                 gestures. Soon NEC, a computer company, was
                                                 using the technology for its airbag controls.
                                                 Once the seat recognizes that it is occupied by a
                                                 small child, the airbags will not inflate, reducing
                                                 the chance of injury or death.

              He gives other examples in which tiny computers could be embedded in anything from
              shelves to name tags to help people with rudimentary tasks. The best part, he said, is that
              people might not even know that the computers were there. "I would rather call it unobtrusive
              computing," he said.

              So far, the kitchen is where most pervasive-computing appliances are emerging. Sunbeam, for
              example, recently announced the creation of a company called Thalia Products (Thalia stands
              loosely for "thinking and linking intelligent appliances) and is planning to offer items like a
              smart coffeemaker, a smart mixer and a smart alarm clock this fall. The mixer will weigh
              ingredients and follow recipes stored on an Internet-connected console, the company's Web
              site says. The clock will synchronize other clocks in the house and give you information
              about other appliances, like the fact that your coffeemaker is out of water.

              Merloni is already selling machines from its new interactive product line, Ariston Digital.

              "Phone the washing machine and practice e-cooking," a recent press release said. The Ariston
              Digital Oven can download recipes from the Web, regulate its temperature based on its
              sensors' information about how the food is cooking and turn itself off when it senses that the
              dish is done.

              The arrival of Merloni's products added urgency to ECCO Design's work for the appliance
              manufacturer (ECCO will not disclose its client's name, but it is not Merloni). Late last fall,
              Mr. Chan and his designers interviewed chefs, parents, anthropologists, consultants and
              software makers. They sent out a questionnaire to 50 people and brainstormed about what
              features consumers would probably reject.

              For example, the notion of a refrigerator that automatically ordered replenishments started to
              sound less feasible. If the refrigerator was supposed to sense that the milk was low, for
              example, the milk would probably have to be placed on the same spot on the same shelf each
              time it was returned so a sensor could track its weight. "But what if a child suddenly puts
              something on Shelf 3 instead of Shelf 4?" Mr. Robertson asked.

              "People fear being locked into a system," Mr. Chan said. If appliances are going to be
              intelligent, he said, they need to be intelligent enough to know when to let human nature run
              its course. At the conference in Maryland, one speaker imagined a day when his trash
              compactor would notice that he had thrown away a box of Wheaties and automatically order a
              new box. But what if he was going to be away for a week or was in the mood for Cheerios?

              Some of the innovations related to smart appliances remind Judy Tso, an anthropologist hired
              by ECCO Design, of the promises of the "smart house," a concept that has been floating
              around among architects and builders for more than a decade. In one version, systems for
              heating, cooling and security adjust automatically, depending on who is home.

              "But the bottom line is that the benefits are wimpy," said Ms.
              Tso, who has her own consulting company, called Aha
              Solutions Unlimited, in Boston. "And it is expensive." With
              any new technology that might be introduced, she said, "there
              is always this tension between what it will cost and what it
              will do for you."

              James M. Utterback, a professor at M.I.T.'s Sloan School of Management, has spent several
              years looking at the life span of technological innovations. If a product is to succeed, price
              plays a critical role, he said. And timing can be everything: Companies that continue to
              manufacture a product when a new standard is starting to evolve, for example, will almost
              inevitably fail, as diskette and typewriter companies have learned too well.

              But one of the main reasons that companies with new products stumble, Professor Utterback
              said, "is that they fail to appreciate or investigate the marketplace." Many companies simply
              ask, "What can we do with the technology?" And once they determine what they can do, he
              said, they assume that people will want it.

              To Professor Utterback, many of the pervasive-computing appliances he has heard about fall
              into that category. "They seem kind of silly to me," he said offhandedly at the end of an
              interview. "Like having your fridge order your groceries. Who really wants it?"
 

The views and opinions expressed in this page are strictly those of the page author.
The contents of this page have not been reviewed or approved by the University of Minnesota.