the hot new medium: paper

From: Eugene Leitl (Eugene.Leitl@lrz.uni-muenchen.de)
Date: Fri Apr 06 2001 - 08:25:09 PDT


http://www.wired.com/wired/archive/9.04/anoto_pr.html

The Hot New Medium: Paper

How the oldest interface in the book is redrawing the map of the
networked world.

By Steve Silberman

They're building something enormous at a research park outside of Lund,
Sweden. Like any concept that eventually becomes the standard by which
imagination is measured, it started out small and grew as its creators
came to understand the scale of what they were making. Now it's half as
big as the United States.

By dawn - which means midmorning in winter at this latitude - the
bulldozers have already been roaring for hours, churning mountains out of
the red, muddy soil, forcing drivers to improvise roads over the fresh
muck. The chaos of construction is not unusual here. The wireless boom in
Europe and Asia is bringing dozens of new office buildings to research
parks like this one all over Scandinavia, from startups incubated as
student projects, to established firms poised to surf the wave of
personal-area-network devices that will wash ashore in the next couple
of years. Even more ambitious, however, is the project taking shape in a
cluttered second-floor office known as the monkey cage by the engineers
and programmers who work there.

As Christer Fåhraeus tries to describe the magnitude of what he and his
team are designing, his fingers dance with a pen across a sheet of paper.
A compact, straw-haired, 35-year-old Swede, Fåhraeus gives off the aura of
a tensely coiled spring. In Swedish or English, he speaks staccato, as if
there were too many ideas backing up behind the frustratingly slow buffer
of syntax. To relieve the pressure inside him, he sketches arrows,
rectangles, and intersections that form the true arc of his thoughts.
Words come secondarily to him, Fåhraeus tells me - his brain thinks in
images. He holds up the paper.

"This is the most advanced digital input screen ever developed," he
declares. "It has very high resolution, perfect contrast, and costs a
fraction of a cent to produce. Any graphical interface can be printed on
it,
and you get years of full-time education, paid for by the government, to
learn how to use it. It will not be beaten in our lifetime."

He puts the paper in my hands. "And I can give it to you, because I have
hundreds more," he offers, gesturing toward a stack of blank paper on
his desk. Fåhraeus isn't handing me a sketch of the input screen. The
paper is the screen.

That's what they're building at Anoto, the company Fåhraeus launched a
year and a half ago beside this muddy field near the southwestern tip of
Sweden: a network that can transform millions of sheets of paper into a
new front end for the Internet.

By the end of this year, Ericsson will bring to market a pudgy-looking
ballpoint called the Chatpen. It will be the first of a new breed of
writing instrument invented by Anoto that will allow you to send email and
faxes directly on paper, with no personal computer or wireless tablet in
sight. You'll be able to jot these messages down on business cards, legal
pads, or company letterhead. To send a message, you'll simply check a box
for "Send as email" or "Send as fax" that's printed in the corner of the
paper. Marking other boxes will route your message to pagers or mobile
phones. A single scribbled note will trigger a cascade of networked
events: Jotting down a lunch date in your day planner could update your
laptop and fire off an email to your assistant.

"There are three fundamental technologies for gathering, storing, and
spreading information - voice, computer, and paper and pen," Fåhraeus
declares, drawing three squares on the paper and methodically checking
them off. "Now we make this one digital and wireless, like the others." By
closing the gap between paper and the digital domain, Anoto is planning
to put the convenience and speed of the Net behind an interface that
was debugged thousands of years ago.

"If we succeed," he adds, "we will have more product coverage than any
other company on earth."

If his network rolls out as scheduled, within a year you'll be able to
make a check mark beside a magazine ad to receive information about a
product, or even to buy it. Visualize ecommerce without the
click-and-wait: Browsing through a printed catalog, you'll purchase items
- software, a subwoofer, or a trip to Paris - by ticking them off with a
pen. Circling your destination on a city map will display, on your
PalmPilot or mobile phone, the quickest route from here to there, movie
showtimes, or tonight's menu at the best bistros in the area.

To do these things, you'll need an instrument like the Chatpen that
contains technology developed by Anoto. (By 2003, other Anoto-enabled
pens, including Pilot rollerballs and a characteristically elegant
offering from Montblanc, will be available.) You'll also need a supply of
the special paper that Anoto has christened digital paper. It won't be
hard to find, and it won't cost much more than standard copy stock. Unlike
Xerox PARC's electronic paper or MIT/E Ink's Immedia, Anoto's technology
employs real paper and commonly available inks. By the time Chatpens
appear in office-supply shops and mail-order catalogs this fall, digital
paper sporting the Anoto logo will be turning up everywhere. This global
rollout will be branded with the most recognizable names in the
office-products industry. You'll be able to buy digital Cambridge legal
pads, digital At-A-Glance organizers, digital Financial Times diaries, and
digital Franklin planners handsomely bound in leather. This winter, expect
flurries of digital 3M Post-it Notes.

At Comdex 2000 last fall, the buzz was that some Swedish startup (or
Japanese; the name of the company, which does sound vaguely Pacific Rim,
is taken from the Latin annoto, meaning "I scribble") had developed a cool
"smart pen." Anoto was one of two finalists for a Best of Show award,
though interestingly, it lost to the Tablet PC, Microsoft's platform that
supports handwriting. Bill Gates' keynote, extolling the virtues of
handwritten input, included a demo in which he beamed a hand-drawn map to
an assistant - with directions to the nearest Starbucks.

While the established heavyweight took home the prize, the concepts
driving the Anoto network are a lot more ambitious than Microsoft's latest
must-buy for the handheld sector. Pen-based interfaces are not exactly
news, even if being able to write in your own handwriting is considerably
easier than trying to recall the Graffiti symbol for the letter q. While
the Chatpen demo in the Ericsson booth - featuring a caricaturist whose
drawings were piped to a laptop screen - was cute, it barely hinted at
Anoto's potential.

Anoto's approach, in contrast to Microsoft's, doesn't require a PC. Each
Anoto pen contains a Bluetooth chip that communicates with any other
Bluetooth device within 30 feet, which could be your mobile phone or
PDA. Ericsson will introduce the R520, the first handset to ship with
Bluetooth, in the US and Europe by the second half of this year. Nokia
already sells a Bluetooth card for its 6210 phone, and Toshiba started
shipping Bluetooth PC cards last fall. The penetration of these devices is
expected to snowball as the cost of Bluetooth chipsets plummets to
between $5 and $10 in the next three years. If the public-access
Bluetooth nodes now in development at companies like Cerulic and
NomadNetworks are widely installed in airports, hotels, and conference
centers, eventually all you'll need to carry is an Anoto-enabled pen and a
sheet of digital paper.

As nifty and convenient as the porting of handwritten text to the Net
may seem, there will be even more advantages to liberation from the
networked typewriter. If your native language is, say, Chinese, Arabic, or
Russian, you will no longer have to translate your thoughts into an
alphabet that a QWERTY keyboard understands. Once symbols and line
drawings are as easy to pour into the datastream as ASCII, you won't
have to depend on text at all. Storyboards, architectural sketches, fabric
designs, game strategies, and comic strips will be emailed, faxed, or
posted to the Web as quickly as they can be sketched by hand. Equations,
with their special characters and sub- and superscripts, will be a breeze.
If you compose a melody on sheet music, you will be able to play it
instantly on your mobile phone or MIDI device. An artist will be able to
zap sketches from his atelier to the Kinko's around the corner.

"Leonardo da Vinci would be our perfect customer," brags Jan
Andersson, the president of Anoto.

     The pen reads a near-invisible grid of gray dots to
     fix a location on a map half the size of the United
     States.

The peerless doodler from Florence might have appreciated that the spark
of genius which makes the network function does not reside in Anoto's
fancy pens. It's printed on the paper.

It's a map.

The first time you see this map, you may not even notice it. Printed in a
shade of carbon-based ink called Anoto Black that's more visible at the
infrared end of the spectrum, the map appears as a light-gray dusting of
dots, forming a nearly invisible grid on the surface of the paper. Each
sheet of digital paper carries only a small portion of the map. If you
look at an Anoto-enabled Post-it, what you're seeing is a Post-it-sized
fraction of a map that is actually 1.8 million square miles - half the
area of the US.

Here's how it works: Beside the ballpoint tip of each Anoto-enabled pen
is a lens. Behind the lens is the same sort of CMOS image-sensing chip
used in cash machines and digital cameras. This tiny, inexpensive eye is
wired to a microprocessor in the pen. Every one-hundredth of a second,
the camera takes a snapshot of whatever portion of the map is
underneath the pen at any moment.

The pen doesn't actually "see" what you're writing. The CMOS chip is
programmed to favor infrared, so the trail of ink is invisible to the
camera. All the pen sees is the map. ("The ink is just there to make you
feel comfortable," purrs a piece of Anoto documentation, with typical
Swedish understatement.)

The dots that make up the map are each one-tenth of a millimeter in
diameter, and they're arrayed on a grid of 2- by 2-millimeter squares, 36
dots to a square. In a basic grid square, these dots would be arranged
smoothly along x/y axes, like perfectly aligned chess pieces. All of the
points on the Anoto map, however, are slightly displaced from those
axes. This displacement creates a unique pattern in each square - and
there are 4,722,366,482,869,645,213,696 squares in all.

To visualize this, imagine that you're writing on a huge sheet of graph
paper. The pattern of dots in any particular square corresponds to an
exact location on the sheet - say, "B2." As you write on the paper, your
pen travels over a series of locations: from B2, to B3, then over to C1 as
you cross a t, and so forth. The movement of your pen over these
locations corresponds exactly to the shape of what you're writing. Using
the map, the pen obtains a precise reading of its position on the vast
grid, down to one-thirtieth of a millimeter. As the instrument dances
across the grid, it stores a series of locations. It time-stamps this
itinerary, in case later verification is needed. There's enough memory in
the pen to store about a hundred notebook-sized pages of writing.

Each instrument is coded with a unique identity. The pen in your pocket
might be number 754348847, for example. When you check off one of the
special function boxes - like "Send as fax" - in the corner of each piece
of paper, the pen transmits the contents of its memory to Anoto-powered
Bluetooth devices in your network. (In addition to the standard Bluetooth
security layer, this information is also encrypted by the pen, using
Public Key Infrastructure and 128-bit keys.) From one of those devices,
such as a phone or PDA with a wireless modem, the burst of information
from the pen hitches a ride to the Net.

It's important to make one thing clear: The Anoto pen does not understand
language. With certain exceptions, the pen doesn't perform OCR (optical
character recognition) on what you've written to translate your
handwriting into standard ASCII text. The email messages you send out
arrive in recipients' in-boxes as small inline graphics files displaying
your words, in your handwriting, exactly the way you wrote it. (When you
send email to a mobile phone, it appears as a graphical SMS message.)

Certain areas on the paper will be dedicated to OCR-related functions.
Letters and numbers - email addresses and subject lines, snail mail
addresses, and phone numbers - written carefully in form fields there will
be converted to ASCII characters. A digital business card, for instance,
might have lines on the back where you can enter contact information.
Checking "Send" would email that information to the person whose name
is on the front of the card.

The pen doesn't know if you've just scribbled a love note, signed a pink
slip, or declared war. It knows only a few things: which pen it is in the
Anoto network, which locations on the map it's been seeing lately, and
what time it is. This information - the pen's identity and a series of
time-stamped locations on the map - is what gets transmitted to the
Net.

There, the encrypted stream of bits from the pen employs standard DNS
protocols, like the ones used by a Web browser, to look for the Anoto Name
Server, or ANS, which is a multimillion-dollar array of Unix drives
currently under construction in Stockholm. The first ANS disk array will
be located there, but eventually Anoto will employ a distributed network
of servers. Each transmission from the pen arrives at the ANS with a
question: "Who owns the part of the map where I am now?"

This is where things really get interesting. The big map is divided into
territories, with a certain number of grid squares allocated for various
products and services. One area of the map, say, might be reserved for
Filofax organizers. Another will be dedicated to Post-it Notes. A third
might belong to a software vendor who runs magazine ads printed on
digital paper. By entering into a partnership with Anoto, you buy your
own little chunk of real estate on the big map, which gives you the right
to print that portion of the map on your products - whether your
business is manufacturing legal pads, booking vacations on cruise ships,
or selling DVDs from a catalog.

By tracking where each Anoto pen is on the map, the ANS knows, for
instance, that pen number 754348847 made a check mark at 12:03 pm in
an area of the grid that belongs to an online florist. The instrument then
transmits a message to the florist's own servers to fulfill the purchase
order, dispatching a dozen roses to the mailing address written at the
bottom of the paper. Because OCR is dodgy business - as anyone who
uses one of the fax-to-email services knows - the shipping address
might appear on the sender's mobile phone for verification before the
bouquet leaves the warehouse.

Anoto isn't in this game to make a branded splash in the paper business,
or to steal market share from Bic. Ericsson's Chatpen will be used to
establish the standard, and the business of designing later versions of
the pen will fall to established players like Pilot and Montblanc. Anoto
won't be making paper, either.

Anoto's business model resides in the ANS - in selling off chunks of the
big map, and tracking every transaction scribbled on digital paper. If a
purchase is made on Mead paper, with a Montblanc pen, and the product
or service is delivered to the customer by an online merchant, then Mead,
Montblanc, the merchant, and Anoto each get a cut. Anoto is betting
that it will be able to tap into revenue streams created by a breed of
paper-based ecommerce services that hasn't even been imagined yet.

Such streams, however, can only run in courses laid by a widely accepted
standard. In that sense, Anoto is a thoroughly 21st-century business: It's
not about making widgets and shipping them to the consumer, or about
building a brand. It's about advocating a standard, and then insinuating a
new enterprise into the global infrastructure created by that standard.

Anoto isn't locked into Bluetooth. If that technology confounds
expectations and is not widely adopted, Anoto could swap the transceiver
out of its pens for anything else that can do the job. However, the chip
inside the pen gives the thousands of companies committed to Bluetooth -
like Ericsson, which is keen on extending its market after facing
unexpected losses early this year - an incentive to adopt Anoto's own
standard for digital paper. Ericsson has picked up 15 percent of Anoto - a
subsidiary of C Technologies, a publicly traded company that is another of
Fåhraeus' ventures - with an option to pick up another 15 percent.

Whether or not Anoto and its partners are able to make every application
come off smoothly - and more broadly, whether the entire project thrives
or stumbles - the network concept offers a preview of the world to come.
At one end of the digital spectrum, Moore's law and fat-pipe upgrades will
deliver heart-stopping clock speeds and insane amounts of bandwidth in the
next few years. But creative innovation is starting to flourish at the
other end, too, where swarms of highly networked and practically
disposable eyes and ears - like Anoto's pens - will be let loose on the
world to swarm around us, listening and testing, and buzzing with the news
of what they find.

A poet once told me that a writer must write with "the mind of God and the
eyes of a spider." Before the Net came along, there was a lot of
speculation that the ultimate product of technology, when we finally had
the hardware to build it, would be something approximating the mind of God
- a centralized superbrain. As we extend our networks into every available
niche of our lives, the ultimate product of technology is turning out to
be more like the sum of the sparks of intelligence in the eyes of billions
of ephemeral spiders, weaving webs in all the corners of creation.

It's at this end of the spectrum that a startup like Anoto could blow
everyone's mind by establishing a standard that is widely adopted, and
scaling upward toward infinity.

With its labyrinth of cobblestone streets, Romanesque cathedral, and
brooding medieval buildings housing bohemian cafés, Lund is the darkly
enchanted European college town of your dreams. An extraordinary number of
shops in Lund bear the sign Antikvariat, meaning "a seller of old books
and antique maps." So it's only fitting that a new kind of map is being
conceived not far from here. The town's fascination with cartography goes
back at least as far as the 14th century, when a fantastically painted
astronomical clock was installed in the Lund cathedral to track the sun
and the moon through the zodiac, marking noon and 3 pm with a blast of
trumpets and a procession of mechanical Wise Men toward the Virgin Mary.
It still keeps good time.

The town, which is much closer to Copenhagen than it is to Stockholm,
lies off the beaten path of the tourist trade, in part because many local
merchants shutter their windows after the students leave for the summer.
Swedes celebrate Lund for the climate of philosophical ferment
surrounding the university, which is balanced by an attitude that Swedish
essayist Jan Mårtensson called the Lund spirit: "an ironic distance to
everything [and] a barb to deflate pompous self-importance."

In 1983, faced with a decline in Sweden's traditional industrial mainstays
- shipbuilding and textile manufacturing - the town broke ground for the
first science park in Sweden: Ideon, where Anoto is located. Ideon's first
tenant was the Ericsson Mobile Telephony Laboratory. The
cross-fertilization of academic research and commercial development -
inspired by US models - proved fruitful. Ericsson's first portable mobile
phone, the C900, was designed at Ideon and brought to market in 1987.

Fåhraeus grew up in Linköping and came to Lund as an undergraduate in
1986. His interests roamed broadly among the sciences, encompassing
medical biophysics, bioengineering, and the mathematical modeling of
neurons. As a teenager, he displayed a knack for making things happen.
When he was 15, he launched the local chapter of a conservative student
organization (the second largest in the country) and founded a reading
circle that received funds from the government. Though Fåhraeus
currently holds 7 patents and has filed applications for 50 more, he says,
"I never thought of myself as an inventor. I never thought of myself as an
entrepreneur. But I was good at taking initiative, inventing games - or
reinventing them if I was losing."

     With the absolute-positioning map, the "interface"
     was no longer the device, but any surface you could
     print the pattern on.

In December 1994, Fåhraeus launched his first company, CellaVision, which
specializes in building microscopy systems that partially automate cell
analysis. It wasn't easy. In the early '90s, the chill winds of a global
recession, stirred up by currency speculation, blew over the Swedish
krona. The Swedish National Bank boosted its lending rate fivefold, while
executives from companies like Volvo floated down on golden parachutes to
countries with lower tax burdens, such as the United Kingdom.

Just as the telecom revolution was getting under way, young Swedish
entrepreneurs scrabbled for sources of venture capital. Fåhraeus
persisted. In June of 1996, he founded C Technologies, which produced
his most successful product to date, a handheld scanner called the C Pen.
(Despite its name, the C Pen, which saw a modest $5 million in sales last
year, is not primarily a writing instrument.) Other products created at C
Tech include intelligent surveillance cameras (spun off to form a company
called WeSpot) and an optical mouse. A year after starting C Tech,
Fåhraeus launched Precise Biometrics, which specializes in fingerprint
recognition.

The common thread running through Fåhraeus' product line is real-time
image processing. What the optical mouse, scanner, and Anoto pen have
in common is that they take digital snapshots of what they see to
determine their whereabouts. The camera in an optical mouse (such as the
Apple Pro Mouse) scrutinizes irregularities on the scrolling surface to
gauge what is called the mouse's relative position. The mouse judges
movement, not location. If you pick up the mouse and set it down, it
forgets where it was and starts tracking from zero again.

The C Pen does let you add your own words to scanned text by tracing
letters on any irregular surface, like a page in a book. But because the C
Pen contains no ink, you don't see the letters you're writing. They're
preserved in the scanner's memory, and you can edit or transmit them like
regular scanned text. As with an optical mouse, however, lifting the C Pen
from its tracking surface cancels out its relative position. To write an
x, you have to use what's called a unistroke - you must cross a single
line over itself without lifting the pen, as you do with a palmtop stylus.

Understandably, this annoyed Petter Ericson, a graduate student at Lund
University who began working for C Tech in March 1997. In a country of
reserved, fair-skinned blonds, Ericson has the swarthy good looks of an
Italian bad-boy film star. Fåhraeus also hired Ericson's high school
friend, Ola Hugosson. While Ericson's vitality seems barely contained
under a thicket of cowlicks, Hugosson is clearly focused inward. A
musician from a family of musicians, Hugosson plays Bach at home on the
piano. He met Ericson when they both enrolled in a national contest for
student programmers. Hugosson was impressed by the lightning rapidity of
Ericson's mind, while Ericson was awed by Hugosson's intensity as he bore
down on a problem. They made a powerful team. Ericson then encouraged
another old friend, Linus Wiebe, to take a job at Precise Biometrics.

All three first learned how to program in the hacker subculture that
sprang up around an illustrious generation of home computers whose
names are legendary among programmers of a certain age: the Apple II,
the Commodore 64, the Atari 800, the Amiga 1000, and the Sinclair ZX
Spectrum. There was a narrow window where 13-year-old
geeks-in-training could sharpen their chops - and make friends all over
the world - by writing algorithms and hacking serious code. Before that,
computers were too expensive. After the introduction of so-called
user-friendly interfaces from Apple and IBM, the good stuff was all
hidden under the hood.

Ericson and Hugosson brought the old hacker brio with them to C Tech,
where they were paid to explore answers to essential questions, and even
write their own operating system. "I used to think, 'Later in our careers,
no one will ever let us do this,'" Ericson recalls. By fall 1997, there
was a stew of applications for real-time image processing simmering there.
One of Fåhraeus' brainstorms involved printing a pattern on a mousepad
that a mouse could then read like a map. It would judge its position
absolutely, without reference to any prior location.

One day in February 1999, after C Tech team members had been mulling
over concepts like this for months, Ericson decided to go home to nearby
Malmö and take a bath.

Like any good citizen in a region of the planet where baking in a
cedar-planked room is considered the acme of relaxation, Ericson jumped
into the tub to clear his mind. "A lot of this work is trying to solve a
problem while you're continuously interrupted by ringing phones and
ICQs. Your mind becomes like a bad Windows system running too many
background processes," he told me geekishly. He's convinced that hot
baths, like saunas, elevate the temperature of the brain, increasing the
flow of uncensored creative thinking.

What came to him in the tub was dots.

Coincidentally, one of the first methods ever proposed for marrying the
ease of writing by hand to the data-crunching velocity of computers also
employed dots. In the '50s, Bell System switchboard operators kept records
of the long distance calls they handled on 2½- by 5-inch scraps of paper
called toll tickets. Operators were scribbling 2 billion tickets a year,
but paying keypunch operators to transfer a year's worth of call records
to punch cards would have cost Bell $32 million. At the Eastern Joint
Computer Conference in 1957, T. L. Dimond of Bell Labs outlined a method
for capturing handwritten data directly. He proposed replacing the tickets
with a plastic tablet energized by a flow of current through the writing
stylus. To standardize the shapes of the handwritten characters so that a
computer could read them, Dimond suggested training the operators to
construct letters and numbers around pairs of dots, a method he called
"dot constraint." Dimond named his device the Stylator.

Most of the pen-based input devices available today use variations on
Dimond's strategy of embedding processing power in the writing surface.
Wacom tablets run magnetic pulses through a grid of embedded wires to
get a fix on the cursor's position. Many digital whiteboards employ
ultrasonic triangulation to do the same thing. Palmtops, of course, allow
users to write on the computer itself.

This was Ericson's flash in the bathtub: A complex pattern of dots would
make an excellent map for absolute positioning. Dots are good "primaries"
for low-overhead image processing, because they look the same no matter
which way you rotate them. Even if the organization of the dots is random,
the device could consult a lookup table to determine which location in the
pattern it is seeing, and thus determine its position. Using such a
table, however, squandered processing time and memory. "This problem,"
Ericson concluded, "has Ola written all over it."

After Ericson told Hugosson about his idea one Friday afternoon,
Hugosson was, indeed, all over the problem. With the many constraints
posed by a system that has to run super-efficiently on a handheld device,
it was a perfect puzzle for a mathematician who could appreciate the
beauty that Bach coaxed out of rigorous symmetries. When Ericson
returned to his desk on Monday morning, he found a sheet of paper, with
a pattern printed on it, waiting for him.

Hugosson had spent the entire weekend pounding out the math,
incubating clouds of dots in C and PostScript. His solution was to
generate the pattern by using an algorithm, rather than depending on
brute processing power to sort through random messes of dots. That
way, the mouse wouldn't have to store the entire pattern in its memory. It
could merely store the algorithm.

The pattern he presented to Ericson consisted of dots in two sizes,
spaced a millimeter apart. With the number of possible permutations of
small and large dots, the size of Hugosson's map was 4 by 4 meters. This
seemed like more than enough area to play with, because the men were
thinking only in terms of a device that would track across a mousepad or
a book.

Then Tomas Edso and Mats-Petter Petterson, two student interns at C
Tech, suggested using dot displacement, rather than dot size, to
determine the unique pattern of locations on the grid. Each dot would
then yield two bits of information, corresponding to the degree of
displacement along the x and y axes, rather than one. Edso and
Petterson's strategy - which employed smaller dots, with 36 per square
instead of 25 - resulted in 272 possible arrangements of dots, rather
than 225. Even printing the dots only .3 millimeters apart - rather than 1
millimeter - and allowing for redundancy to correct for scanning errors,
the size of the grid became much larger. "How big is the pattern today ?"
members of the C Tech team would tease one another.

From an area that would have covered 10 letter-sized sheets of paper,
and then several football fields, the pattern eventually grew as big as
Belgium, and finally to its current dimensions: the equivalent of
73,000,000,000 letter-sized sheets, or the area from the Pacific Ocean
to the Mississippi, and from the Rio Grande to the Canadian border.

Clearly, there was a lot of there there. But what to do with it?

One day in the lab, Ericson smashed open a Bic pen, extracted the tip
and ink supply, and taped it to a C Pen. Suddenly the "interface" was no
longer the device itself, but any surface that you could print the pattern
on.

"We realized we were sitting on something fundamental," Fåhraeus told
me. "We were on a clean sheet of paper."

It was Fåhraeus who had the bold notion of seeing that virtual territory
as marketable real estate in an enormous network of partnerships. For the
last six months, Örjan Johansson, chair of the Anoto board, has been
pounding the pavement, negotiating partnerships with 3M, Mead,
Ashford, At-A-Glance, Charles Letts, Time Manager International,
Filofax, and Time/System International in the US, as well as Esselte,
Unipapel, and Hermelin in Europe and Kokuyo in Japan - each market
leaders in their region. These companies have everything to gain: The
technical demands and costs of printing the Anoto pattern on their
products are trivial. (You can do it with any printing system that has
1,000 dpi resolution, and Anoto is developing plug-ins for Quark and
PageMaker as part of its developer's kit, which should be available by the
time you read this.)

"When you tell these companies that they can go from manufacturing plain
paper to becoming a service provider with paper as the portal, they
realize they're in a whole new ball game," Johansson told me. By
convincing the largest manufacturers in every paper market to print its
dots on their products, Anoto is establishing a de facto global standard
before consumers have even heard of the company.

At Ideon, Fåhraeus sketched out Anoto's three-year plan for world
domination. First: License the pattern and pen technology to Anoto's
initial partners for a song - in some cases for free - to establish and
disseminate the standard. Then, in year one, sign partnerships with mobile
phone companies and telcos, who see the technology as a way to siphon
a flood of ecommerce sales through their own billing channels. A Chatpen
would be a natural accessory to sell, or give, to the owner of a Bluetooth
phone; the company that bills for the phone service could then take a
slice of every scribbled transaction. Sonera, the second-largest telecom
provider in Scandinavia, signed on in February, and announcements of
other major providers are expected at this year's CeBIT. When you sign
up for Sonera's mobile service, you'll be offered the option to subscribe
to Anoto's paper-based services.

Year two: Market the technology in consumer-electronics stores,
branded with names now associated with personal computers. Year three:
Imagine racks of Anoto-enabled Parker pens for sale at Office Depot,
beside shelves of digital paper.

     People will link their beloved paper to the flow of
     digitized information, and paper-based ecommerce
     is just the beginning.

In January 2000, Linus Wiebe left Precise Biometrics to join Anoto as
director of new concepts. Now his job is to weigh the business potential
of applications such as putting the Anoto map on standardized tests,
creating real-time lotto games, transmitting prescriptions directly to
pharmacies, printing active hyperlinks in books, and even laying the
pattern over the floors of warehouses so that forklifts can be steered
remotely.

As descriptions of Anoto's technology have started to spread around the
world, the proposals that land daily in Wiebe's in-box reach farther and
farther afield. One email he received from a university proposed printing
the pattern on a curved semi-transparent plastic surface and embedding
the optics, processor, and Bluetooth transceiver into a 2- by
2-centimeter cube. "I still do not know the problem they were trying to
address," Wiebe observes dryly.

For Hugosson, this flood of potential applications for his algorithm is
immensely gratifying. "When you can invent an idea that is so basic that
it leads to other ideas," he says, "an idea that seems to have many
implications, an idea that leads to a completely new set of thinking ..."

He's quiet for a moment. "I don't know how to express it in words," he
concludes.

Why would a posse of hotshot Swedish coders - or more to the point, a
struggling telecom giant like Ericsson - hitch their wireless wagons to a
fading star? Everyone knows the paper industry must have been hit hard
by the overnight ubiquity of the PalmPilot and the Handspring Visor. And
surely, with Bluetooth - or a couple of years out, with 3G - the
2,000-year-long Age of Paper will soon be over.

So whatever happened to the paperless office? There's actually more
paper in the digital office - cascading out of printers and clogging up
copy machines - than there was 10 years ago (30 percent more,
according to a report by the American Forest and Paper Association).
Like the paperback novel that was supposed to be supplanted by every
technical marvel from the radio to the ebook, paper solves more problems
than it creates. We don't hate paper - the way we hate insipid broadcast
TV, tangled telephone cords, futzing with Wite-Out, five-day waits for a
letter from across town, and stores that are open only at certain times of
day. It's the inert state of the data stored on paper that is a vexing
anachronism, and this is precisely the problem that Anoto addresses.

Palmtops are indeed eating into the market for a specific kind of product:
the kind of paper - in day planners, calendars, and little black books -
that stores data that wants to communicate with our digital networks.
Executives for Mead and Franklin Covey carefully acknowledge that there's
been "a flattening" of the market in the last two years, specifically for
formats that handle contact lists and appointment schedules.

The repeat business in this sector is still the envy of other industries.
Many of the 50 million customers who buy a couple of At-A-Glance calendars
every year, to say nothing of the estimated 800 million who buy diaries,
will part with their tattered paper products only when they're pried from
their cold, dead, pulp-loving fingers. Those buyers are so fanatically
loyal to specific model numbers and form factors that John Hayek, a senior
VP of marketing for Mead, calls it a "religious" market. Anoto will
furnish a way for people to link their beloved paper to the flow of
digitized information. For Mead, Hayek says, betting on Anoto is "an
acknowledgement that a paper-based organization is looking forward in the
21st century."

One area of the office market that is exploding, rather than flattening,
is in products that make it easier to "co-use" PDAs and paper-based media.
For the past few years, the companies Anoto is courting have been
struggling to find their footing in a changing landscape where critical
information is segregated on both sides of the paper-versus-digital
divide. "Right now, we're all trying to synchronize Palms, phones,
Outlook, day planners, Web sites, and thousands of floating Post-it
Notes," says Jeff Anderson, VP of Franklin Covey's Eproducts and Planner
division. "It's almost seamless now, but the big glaring gap is
paper-back-to-digital. Anoto is the last leg to the full solution."

Despite all the clicking on the Web, there are many areas of the paper
domain where digital has barely made a dent. The annual cost of
processing forms that are filled in by hand, such as tax returns, is still
$700 billion in the US alone. There are innumerable applications in which
digital input could migrate happily to paper: imagine putting a check mark
in the newspaper to program your VCR. Anoto pens themselves will be
customized by users who check off options in manuals, the interface
fine-tuned by - what else? - ticking boxes on paper. If you want to
customize the color of the ink or the texture of a line in an email
message,
for instance, you'll choose from a list of options on a printed menu. Or
you could create a virtual flip book by sketching a series of drawings and
selecting a box labeled "Send as GIF animation."

"From a paper-product manufacturer's perspective," says Hayek, "we're
talking about a little more ink for the dots. It's an easy do."

Like the other startups rising out of the mud at Ideon, Anoto represents
not only a new generation of aggressively innovative Swedish IT
ventures, but a new generation of Swedes.

At a restaurant with a sweeping view of Stockholm's Old City, I washed
down eight kinds of herring with five flavors of akavit with Per Bill, a
member of parliament who championed many of the changes in policy that
have made his country a leader in the emerging wireless landscape. He
articulated the changes in the Swedish psyche wrought by the economic
upheavals of the last decade.

"In California, it's fun to be rich, and it's also OK to go bankrupt. In
Japan, if you fail, you're supposed to commit seppuku," he said. "Ten
years ago in Sweden, you were not allowed to be rich, and you were also
not allowed to fail."

Bill used a virtually untranslatable word as the key to understanding the
avoidance of both conspicuous success and humiliating failure that is
deeply ingrained in the Swedish psyche: lagom. Meaning something like
"lukewarm" and "just enough," lagom is the inclination among Swedes to
shun ostentation, accept modest rewards, be good team players, and fly
under the radar. In the past, the positive side of lagom caused Swedes to
refuse to allow the kind of desperate poverty in their ranks that is
business as usual in the US. The negative side could be seen in the
tendency for the Swedish government to expend energy twiddling with
the economy from the top down, rather than seeding it from the ground
up by providing incentives for entrepreneurs.

What shook up the lagom mindset, he explained, was the recession.
Before 1994, Swedes thought that 4 percent unemployment was a
catastrophe; within a year or two, more than one in ten Swedes found
themselves out of work. A precipitous change in economic status became
less of a cause for loss of face. To compete in the global IT marketplace,
Bill observed, "you have to embrace the possibility of failure."

Startups like Anoto are training grounds for a generation of Swedes who
grew up watching Dallas and Falcon Crest, becoming early adopters of
technology, and absorbing American notions of ambition. Now that
they're launching their own companies, says Lars-Fredric Hansson of
Ernst & Young ePartners, they've been getting assistance from an
unexpected source. The managers who made golden-parachute exits to
countries with lighter tax burdens when the recession hit are coming
home to Sweden to invest in the next generation of IT entrepreneurs.

There's a fitting symbol of this convergence of old and new economies
outside the Ericsson headquarters at Ideon: a stone monument to Harald
Bluetooth, who was the king of Denmark in the 10th century. In one hand,
he holds a phone, and in the other, a laptop. The significance of the name
Bluetooth for the group that developed the wireless standard originates
with the king, who threw parties for his fiercest rivals. These wild
affairs would "help clear out the bad feeling," said board chair
Johansson, who came to Anoto after spearheading the Bluetooth initiative
at Ericsson. Many members of the original Bluetooth coalition had long
been cutthroat competitors: Ericsson and Nokia; IBM and Toshiba; the US
versus Japan.

"My aunt used to hold out her closed fist and say, 'How much can you get
in this hand? It's much easier to get something in this hand,'" Johansson
explained, relaxing his fist. In this way, the monument stands as a symbol
of the positive side of lagom - knowing when to lay aside your arms,
even in the presence of your rivals, to achieve a common goal.

That's why Johansson wasn't particularly disappointed that Anoto lost to
Microsoft at Comdex. When the Anoto team saw the Tablet PC, he says,
they immediately began thinking of ways to build bridges between the two
technologies. "Maybe they're creating the back end, and we're creating
the front end. Writing on glass - how fun is that?"

Not as much fun, Anoto is betting, as making it possible for the mind of
the Net to reach as close as the page of the magazine you're reading right
now.

"The paper companies have been feeling that the digital train had passed
them by," Wiebe told me in a Stockholm coffee shop. "We're throwing out
a hook, saying, 'C'mon, join us. Paper has been right all along. We're
bringing you with us.'"



This archive was generated by hypermail 2b29 : Sun Apr 29 2001 - 20:25:33 PDT