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?"