Unlike athletes whose astronomical salaries form huge portions of the costs of operating a professional sports franchise, the team at Xerox Corp.’s Californian-based Palo Alto Research Center (PARC) are expected to achieve big results with relatively small resources.
According to PARC director John Seely Brown, the lab, which is the most famous of Xerox’ four research centres, is only budgeted to receive one-third of one per cent of the company’s total revenue.
“To everybody in the corporation except the CFO, we represent a round-off area.”
In contrast to its financial status, Seely Brown said the PARC team is counted on to be star performers.
“Our job is basically to hit the home runs. And if you also have the bases loaded when you hit those home runs you win a few more games.
“Our job is to be out on the fringe — to let the emergent happen. We are keenly aware of what is going on outside. We think of ourselves as almost the genetic variance that helps the species evolve. We’re in the spirit of looking for the high risk and the ultra high value payoff.”
The staff at the research centre, which created and then lost control of some of the most important advances in computing — including the GUI, Ethernet, WYSIWYG editing, bit-mapped displays, distributed computing and client/server architecture, optical storage and the mouse — is hoping at least one of its crop of rookie ideas will evolve into the next game-winner.
Of all the ideas that the PARC crew are willing to talk about, digital paper is the sexiest and easiest to market.
“Think of it as the world’s first truly erasable paper,” Seely Brown said. “You can print the day’s news on yesterday’s paper. The paper recycles itself.”
Seely Brown said Xerox is finally solving some of the difficulties of getting digital paper to the commercial market. The “paper” is composed of thousands of randomly distributed bichromal microscopic balls. The balls measure about 100 microns in diameter — approximately the same as a human hair — and each is half black and half white, and is wrapped in a transparent rubber coating. The two sides carry opposite charges.
Writing on the paper is done when it is exposed to an electrical stimulus which sets the balls in place, with either their white side or their black side showing in order to produce the desired image. Reusing the paper requires subjecting the sheet to another charge. The rubber coating acts as a shield keeping the balls, which, depending upon the PARC official speaking, are either made of plastic or plant-based materials, safe from random electrical fields.
To date, Seely explained that producing the paper in large quantities has proved difficult and expensive, but that is being resolved. Recently Xerox took possession of the first large roll to be manufactured.
According to Matthew Howard, member of the PARC research staff and a development engineer, the world is likely to see its first commercial glimpse of digital paper within 36 months.
“The compelling market will be large area displays — poster-sized signage with low information content.”
Howard added that PARC researchers are currently trying to reduce the size of the balls in order to increase the paper’s resolution. He also said the researchers are currently working on ideas and patents that could possibly lead to full colour digital paper.
Another invention causing some excitement at PARC is the first continuously operating blue laser, which produces fine dots more quickly than conventional red and infrared lasers.
“Blue light lasers get the resolution higher,” explained Ross Bringans, laboratory manager, electronic materials laboratory.
“The optical system in a laser printer is what they call defraction limited, which means that the resolution of the image depends on the wavelength. Also, the wavelength for an infrared laser is a two-times greater factor than the wavelength of a blue laser, so in principal you can take an infrared laser printer and exchange it for a blue laser and get twice the resolution.”
The Xerox contribution to blue laser research — developing the technology to make it a continuous beam — is important because printing and optical recording require a constant and not intermittent pulse of light.
Despite that development, blue lasers are still not ready for commercial use.
“It turns out that as you go from red to infrared to blue, it becomes more and more difficult, so the efficiency of the laser tends to be less. But we believe that with the rate the technology is improving it will become easily usable,” Bringans said.
Data Glyphs and Intelligent staples
“Think about digital copiers,” instructed Seely Brown. “What you are doing now is scanning in a document, capturing it, and doing information processing — scan, think, print.
“So why not scan, think, link — to the world’s networks — then print? (We will) eventually copy to the Web as easily as you copy to paper. Why not do archiving copying where everything you copy gets digitized and stored when you copy? What a beautiful way to bridge the virtual and physical worlds. Maybe it is even possible to view this as a portal to the World Wide Web and pull documents back out.”
To that end, PARC is developing a couple of technologies. For starters, Xerox is working on intelligent staples, a technology which comes from the world of agriculture. Companies are using intelligent matter as a way of tagging cows with identification markers, and pet owners can also tag their cats or dogs.
Seely Brown explained that the matter (or tags) contain information which can be read by a scanner. That way when his cat gets lost, all somebody has to do is scan its tag and the cat’s Web page — containing his owner’s information — pops up.
In the office environment, staples are created out of the intelligent matter. They have no internal power source and absorb energy from the scanner’s radio frequency transmission. Encoded into each staple is information about the document it secures.
“The staple contains the URL of the document,” said Seely Brown, “that way you can just print the summary of the document, and when you want the most updated version, you just wave the document in front of the wand [or scanner] and up comes the most recent version of the document.”
And just like the staple, the summary document is also more than it appears to be.
First, the summary was probably never even compiled by a human. Instead, technology from the Xerox affiliate Inxight Software Inc. analysed the entire document for word patterns and language use and then created the summary based on that analysis, plus phrases and sentences found within the original document.
Also, the summary is more than likely marked with a data glyph, a small design of self-clocking error correcting code that works like a steroid enhanced bar code.
“Data glyphs have the function of barcodes but with much higher information density. And they look prettier,” explained Dr. Per-Kristian Halvorsen, PARC laboratory manager and principal scientist.
The glyphs, which are 24 bytes (eight bytes being randomly generated numbers and 16 bytes comprising a unique identifier) can be encoded for security and can contain the entire document summary plus hyperlinks. They are scanned into a Xerox Docucenter printer/copier as TIFs and are transported across networks using FTP.
According to Leigh Klotz, of PARC’s document portal business unit, “when documents have data glyphs the paper becomes something more than paper, it becomes something active. The document is a token that stands for a document stored on the network. You can print out a Web page on paper, but the check boxes on the paper become live hypertext links, and the print-out can also be converted into an e-mail or a fax without rekeying it.”
The world has seen numerous changes involving how printed and photocopied words and images find their way onto paper, but it has yet to witness a revolution in how the paper gets moved through printers and photocopiers. Sheets jammed in between inaccessible rollers seem to be an eternal pain, but that might change if the PARC engineers are able to perfect a method that moves paper through a system using nothing but air.
“We are working on a system that is based on air jets,” said Bringans, “that will switch on and off in order to move and rotate the paper….and the hydrodynamics will keep the paper in place, even if [the unit] is turned upside down.”
He said this type of technology “requires a great deal of computation,” because it requires finite control over a large number of tiny air jets, miniaturized to fit onto circuit board technology.
“You have to look at control issues,” Bringans said. He explained that as a piece of paper moves along, the jets under the centre of the paper can be treated and controlled as a group, but the jets at the edges of the paper – the ones that are responsible for directing movement and direction — must be controlled individually in order to create precision movement.
Even if the technology comes to nought for transporting paper through printers, Bringans said the studies in miniaturization and control will prove to be a valuable experience.
Another experience that the PARC group is finding valuable is studying the Internet itself. As one of the companies who worked on the newly accepted IPv.6.0 protocol standard and as an organization working to develop HTTP-NG – a next generation transfer protocol that is friendlier to images – Xerox wanted to know how people use the ‘net. To that end, Halvorsen said he “pulled all of the Web’s content into one lab and placed it on a bunch of servers.”
That massive download, which comprised only the text found on the Web and not the graphics, amounted to just less than 500GB of data (as of mid 1998). He added that downloading Web images would only increase the storage size by a factor of ten. After performing that task, Halvorsen said one of the findings is that people don’t tend to stay on Web sites past three clicks, so designers had better learn to give people what they want more quickly.