Slashdot | Why Kindle 2's Screen Took 12 Years and $150 Million
Paul D. Fernhout
http://hardware.slashdot.org/article.pl?sid=09/02/26/1857251
"Critics are eating up everything about Amazon's Kindle 2 e-book reader
except its $359 price tag. But if you think that's expensive, take a look
behind the Kindle at E Ink, the Cambridge, MA, company that has spent $150
million since 1997 developing the electronic paper display that is the
Kindle's coolest feature. In the company's first interview since the Kindle
2 came out, E Ink CEO Russ Wilcox says it took far longer than expected to
make the microcapsule-based e-paper film not only legible, but durable and
manufacturable. Now that the Kindle 2 is finally getting readers to take
e-books seriously, however, Wilcox says he sees a profitable future in which
many book, magazine, and newspaper publishers will turn to e-paper, if only
to save money on printing and delivery. (Silicon Alley Insider recently
calculated that the New York Times could save more than $300 million a year
by shutting down its presses and buying every subscriber a Kindle). 'What
we've got here is a technology that could be saving the world $80 billion a
year,' Wilcox says."
http://www.xconomy.com/boston/2009/02/26/kindling-a-revolution-e-inks-russ-wilcox-on-e-paper-amazon-and-the-future-of-publishing/
http://www.businessinsider.com/2009/1/printing-the-nyt-costs-twice-as-much-as-sending-every-subscriber-a-free-kindle
I posted this just for what it says in the last two sentences -- that buying
every subscriber a Kindle is cheaper than printing the NYTimes for them for
one year. That's an amazing change. And the electronic version will only get
cheaper and better from here on in. That was one of the big points of the
OLPC project most people had trouble understanding -- whatever the OLPC
could do or could not do as a computer, it would save a lot of money on
textbook distribution even in materially poor countries where the books were
cheaper. And in that context, of saving the world $80 billion a year in
printing costs, the R&D costs of $150 million for the electronic paper are
trivial. But then why was it so hard for the world to get together and
invent it? Why is it so hard for the world to invest in improving its most
critical infrastructure?
Granted, it is still easier and less stressful to carry a newspaper and
magazine and leave it somewhere than to safeguard a portable computing device.
Actually, the electronic ink ideas have been around for some time, but it
has taken a long time, as mentioned above, to make them practical. CMU since
the late 1980s had people pushing for high density screens and electronic
books, only recently are we getting more affordable high density displays.
In the 1980s at a student environmental conference I proposed being able to
run printed paper back through a photocopier or laser printer and have it
change what was on the page. I still like that idea -- to have a cheap but
reuseable media that acts like paper and takes another device to configure
it somehow. Computing devices are still more brittle and fragile than flimsy
paper-like stuff.
Anyway, from the xconomy link, Russ Wilcox on bringing ideas into
production: "I co-founded E Ink with three fellows out of MIT and with Jerry
Rubin, the founder of Lexis-Nexis. I wrote the business plan in my study,
and got copies bound at Staples, and mailed it out through Kinko’s, and all
that. I did all the things you should apocryphally do when you’re an
entrepreneur. At the time, we had no idea it was going to take so long. It
may be that naivete is your friend when you’re starting out in such a
daunting venture. We understood that it was probably going to take two years
to make something that people wanted to buy. And in terms of making
something that looked good, we did that. But what we didn’t see in the
beginning, and learned over time, was that it would take another two years
to go from something that looked good to something that would look good for
many years under all operating conditions—in other words, to achieve
stability and robustness. And then it would take another two years to get
something that you could reproducibly manufacture, at an affordable cost point."
Could open manufacturing have changed that R&D process somehow?
By the way, from the second page of that article: "But if you want to buy a
development kit and design your own device, it’s $3,000. With the
development kit, you get everything that’s inside an e-book, including a
little chip that runs Linux and a bunch of open source drivers, a touch
screen with pen input, and the Broadsheet chip. People are doing all sorts
of fancy stuff with that. There’s a fellow in Malaysia who has ported the
Linux X Windows system to the device, and there are some folks on the West
Coast who have ported Android to it. So part of what we’re doing is just
making this open and trying to let lots of people find interesting uses for it."
Also: "And in the very long term, it should be cheaper than LCDs, because
what you can do with E Ink that you can’t do with LCDs is manufacture them
using a roll-to-roll process. If you ever want to make a billion of anything
cheaply, you print it.We announced last year that we are working with
Hewlett-Packard. They have set up a process to develop roll-to-roll, impact
lithography printing of active-matrix panels. So they can print a backplane,
and our stuff comes in on top. That’s five, seven, maybe 10 years away. So
it’s clearly not tomorrow. But in the long term, E Ink should be very
competitive on price."
So, Russ Wilcox suggests printing is the way to make things cheaply, even
when those things are electronic items replacing printed paper. :-) But it
is not the one-off printing like RepRap; it is the production line printing
of roll-to-roll with layers.
--Paul Fernhout
ben lipkowitz
> In the 1980s at a student environmental conference I proposed being able to
> run printed paper back through a photocopier or laser printer and have it
> change what was on the page. I still like that idea -- to have a cheap but
> reuseable media that acts like paper and takes another device to configure
> it somehow. Computing devices are still more brittle and fragile than flimsy
> paper-like stuff.
you could do this with e-ink (microencapsulated latex balls printed with
charged ink) by passing a row of electrodes across it on either side, or
more likely by passing the paper through a thin "printer" which would save
tremendously on the cost of integrating electrodes into the paper. I'm
picturing something either like a paper laminator or like the
"magnadoodle" toy which has a magnet bar you sweep across the paper to
erase it, except instead of erasing it can draw pictures and text. Not
quite as convenient as pressing a button, but it could be worth the effort
for reduced expense and fragility.
where do I get samples of e-ink paper? (ten years later and this is still
a problem? come on now)
>
> Anyway, from the xconomy link, Russ Wilcox on bringing ideas into
> production: "I co-founded E Ink with three fellows out of MIT and with Jerry
> Rubin, the founder of Lexis-Nexis. I wrote the business plan in my study,
> and got copies bound at Staples, and mailed it out through Kinko’s, and all
> that. I did all the things you should apocryphally do when you’re an
> entrepreneur. At the time, we had no idea it was going to take so long. It
> may be that naivete is your friend when you’re starting out in such a
> daunting venture. We understood that it was probably going to take two years
> to make something that people wanted to buy. And in terms of making
> something that looked good, we did that. But what we didn’t see in the
> beginning, and learned over time, was that it would take another two years
> to go from something that looked good to something that would look good for
> many years under all operating conditions—in other words, to achieve
> stability and robustness. And then it would take another two years to get
> something that you could reproducibly manufacture, at an affordable cost point."
>
> Could open manufacturing have changed that R&D process somehow?
there was what I consider a perfectly functional system as early as 2005:
http://www.positron.org/projects/A51/
and slightly lesser functionality in 2003:
http://www.positron.org/projects/juicebox/
instead we had to wait around until big corporations like sony and amazon
perfected their digital restrictions management software.
> By the way, from the second page of that article: "But if you want to buy a
> development kit and design your own device, it’s $3,000. With the
> development kit, you get everything that’s inside an e-book, including a
> little chip that runs Linux and a bunch of open source drivers, a touch
> screen with pen input, and the Broadsheet chip. People are doing all sorts
> of fancy stuff with that. There’s a fellow in Malaysia who has ported the
> Linux X Windows system to the device, and there are some folks on the West
> Coast who have ported Android to it. So part of what we’re doing is just
> making this open and trying to let lots of people find interesting uses for it."
All I want is the paper itself..
> Also: "And in the very long term, it should be cheaper than LCDs, because
> what you can do with E Ink that you can’t do with LCDs is manufacture them
> using a roll-to-roll process. If you ever want to make a billion of anything
speaking of rolling.. if you don't have a glass panel with ITO conductors
in it, you can roll the paper up to fit in a small case.
> cheaply, you print it.We announced last year that we are working with
> Hewlett-Packard. They have set up a process to develop roll-to-roll, impact
> lithography printing of active-matrix panels. So they can print a backplane,
> and our stuff comes in on top. That’s five, seven, maybe 10 years away. So
> it’s clearly not tomorrow. But in the long term, E Ink should be very
> competitive on price."
>
> So, Russ Wilcox suggests printing is the way to make things cheaply, even
> when those things are electronic items replacing printed paper. :-) But it
> is not the one-off printing like RepRap; it is the production line printing
> of roll-to-roll with layers.
even automation can be automated :)