"A hanging garden of Babylon overlooking Silicon Valley," is Mr Rich Gold's whimsical description of the most famous research laboratory in the history of computing. Set into the hills above Stanford University, the blocky buildings of Xerox's Palo Alto Research Centre, or PARC, have tiered gardens, roomy windowed offices, and views down on San Francisco Bay and the roofs of the corporate powerhouses that worship the silicon chip.
Oddly, the place where virtually all of modern computing was invented is not a computer company at all, but a self-proclaimed "document company". Windows, screen icons, the mouse, the laser printer, WYSIWYG (what you see is what you get) printing and Ethernet (a method of connecting computers into a high-speed local network), even funny screensavers - all were developed in the 1960s and 1970s in Xerox's PARC. At the time, the lab had gathered one of the finest teams of computer-minded researchers together and let them explore their interests in an extremely open, creative environment.
Famously, Xerox brought few of these ideas to market successfully. That task was left to Mr Steve Jobs, co-founder of Apple computer, who convinced PARC to give him a demonstration of its technologies, then promptly worked nearly all of them into Apple's ground-breaking, user friendly Macintosh computer in 1984. Controversy still exists over whether Xerox's ideas were borrowed, stolen or willingly exchanged - the protagonists all have differing stories.
That bit of computing lore forever sets the context for any discussion of the lab and still pains the people at PARC. Mr Gold, manager of PARC's Research in Experimental Documents group, is anxious to point out, for example, that Xerox did bring out an early PC, the Star, which incorporated some of Xerox's research.
But the Star was far more expensive than competitor machines and sank. Unfortunately for Xerox, the Star remains an obscure footnote in computing history (its non-commercial prototype, the Alto, is better known to computer buffs).
Such twists and turns of fate leave Xerox working hard these days to re-establish its digital credentials since the company hopes to forge a new reputation as an Internet technologies and business solutions company. Since the appointment last December of a new head of research and technology, Mr Herve Gallaire, Xerox has stressed its e-business-oriented research.
Regearing the company towards this personality change has in part accounted for Xerox's poor quarterly results recently, says Xerox president and CEO Mr Rick Thoman.
But because of its long and respected history in the youthful technology industry, PARC has a glow that is largely undimmed by something as ephemeral as quarterly results. "PARC is a cauldron of ideas, of thought, of new technology," says the ponytailed, outgoing Mr Gold, whose official company biography also describes him as an "applied cartoonist" and "former consultant in virtual reality". PARC "is like a great university, with all the kids taken out. We actually eat lunch together. We create new ideas together."
Ideas don't always become funded research, and the process of deciding what ideas might produce an actual product or advance knowledge usefully remains one of PARC's greatest challenges, says Mr Gallaire. "You have a large number of ideas that you need to select from," he says. "Selection at an early stage of an idea is extremely difficult."
Mindful of the past, when senior executives astonishingly dismissed most of the work on the Alto, Xerox now has "organisations and processes that evaluate ideas to make sure we realise the full value of [them]."
One idea that initially seemed to have little commercial value has just become the centrepiece of a Xerox spin-off company called ContentGuard.
Launched two weeks ago at PARC, ContentGuard will create and market software that protects and manages copyright content on the Web - limiting the number of times a person can listen to a downloaded song, for example. Microsoft holds a minority share in the company and intends to make ContentGuard central to its own products.
Somewhat more outlandish is digital paper, a product nearly ready for commercial launch - once Xerox figures out a business model for it, says Mr Gold. Composed of sandgrain-sized two-colour balls embedded in a flexible plastic sheet, digital paper is a reusable medium that is "printed" when an electric current passes over the paper. This causes the balls to show one colour or the other, forming letters against a background. Xerox expects it to be used initially for large, easily-changeable signs in department stores, but predicts it may one day form a single-sheet daily paper on which each new page would generate itself.
Even more curious are prototype modular robots that can move by themselves, link and modify their function as part of a larger unit of modules. Such "smart matter" now makes simple snake and spider forms but researchers want to create them on a microscopic level, where millions would gather and create working objects. "You could buy a box of modules and tell them to make a printer," suggests Mr Gold, although he adds dryly that this particular research area "won't be a product soon".
Miniaturisation is a special area of interest at the lab, which has a full-blown MEMS (micro electro mechanical systems) division. MEMS are minute machines built on a silicon chip; researchers have created, for example, tiny mechanised tweezers capable of grabbing a single bacterium. Xerox is interested in optical MEMS and the ways they could enhance printers and copiers. Now such devices run on a single or dual laser focused through a mirror - a printing mechanism that is currently book-sized - but MEMS could make them pea-sized.
One ongoing project involves placing hundreds of silicon mirrors on a single microchip, then tilting them to direct an array of several hundred lasers.
Another project aims to place thousands of lasers on a single chip - Xerox showed off a 1cmsquare chip that packs in 10,000 lasers, the densest laser array in existence. These technologies would enable what researcher Mr Eric Peeters calls "bat out of hell printing" - extremely fast, high quality printing.
Yet another division is exploring the use of multiple microchip sensors that could eventually make inanimate matter intelligent. Researcher Mr Feng Zhao says, for instance, that peppercorn-sized sensors could one day be sprayed across a road surface, enabling the road to tell a car the traffic conditions for a region and select the best route. He imagines a world of smart environments in which sensors would provide factual information that would be analysed by computers, enabling humans "to bridge two worlds - the physical world and the digital and virtual world," he says.
One practical application is copier parts that self-diagnose and alert technicians when they are going to fail.
Underlying all Xerox research is ethnography - the study of human behaviour, how communities and individuals interact and convey knowledge. This is perhaps the oddest element of the lab, the fact that Xerox employs not just scientists but social scientists - some 15 ethnographers - across its cluster of labs. Xerox ethnographer Mr Jack Whalen says ethnography is "at the very heart of this centre" because of a study PARC did decades ago, when researchers videotaped a couple of people who were told to make certain kinds of copies on a copier they had never used before.
The engineers and researchers were stunned to see that real people didn't interact with the machine in the ways they expected and were confused by features that were supposed to be self-explanatory. The video led to a radical rethink at PARC and now ethnographers routinely work on all research projects with the scientists.
"Truly useful technology supports and enhances natural human capacities and practices," says Mr Whalen. "We're trying to look five, 10, 20 years down the road and ethnography helps with that human-scape envisioneering."
All of which may sound rather bizarre, but out of the strange and idiosyncratic often come the flashes of brilliance and breakthroughs that mark a pure research laboratory. "PARC is about inventing the future," says Mr Whalen.
"We're not inventing the next copier here."
klillington@irish-times.ie
