YORKTOWN HEIGHTS, N.Y. -- On a bright, warm morning at IBM Corp.'s research center here, seven of Big Blue's scientists gathered around a conference table to consider a nonscientific question: What helps inventors invent things?
IBM brass had asked the researchers to design a class that could teach lab managers how to help inventors stay fresh and innovative. Quickly the group erupted with ideas for the class's title, its methods, even whether someone could fail it.
Then a boyish-looking operating system programmer, Michal Ostrowski, wondered aloud if the group had made some false assumptions.
"Is it innovation if everyone can see that it is?" he asked, drawing a few murmurs of agreement. "Innovation is not obvious at the time."
Such scientific soul-searching pervaded IBM's inaugural "Innovation Days," a weeklong stretch in September when the technology giant asked 3,000 researchers at eight labs around the world to take off their goggles and re-examine their jobs.
IBM let The Associated Press observe several events, providing an inside look at a conundrum facing many technology companies: how to keep their researchers creative while also demanding they produce short-term results on targeted projects.
The brainstorming sessions about nurturing creativity were just part of Innovation Days.
Individual labs - three in the United States and ones in India, Israel, Switzerland, China and Japan - also created their own programs aimed at recharging everyone's batteries. Researchers practiced tai chi and yoga together. Others ran in a five-kilometer race or played music.
Speakers gave pep talks about creativity - including a master chef, a NASA astrophysicist, technology investor Esther Dyson, a newspaper cartoonist in India and a fragrance chemist.
"Involve customers with labs more often," John Wright, business development manager for International Fragrances and Flavors Inc., told a group at IBM's Thomas J. Watson Research Center in Yorktown Heights. "It's time-consuming, and only some customers want to do it, but it works."
Technology companies used to let scientists do pretty much whatever they wanted, comfortable knowing that eventually, they would come up with something that would bring the company patents, products or prestige. That freedom helped scientists at Bell Laboratories, IBM and other companies win Nobel Prizes.
But sometimes researchers strayed too far from the company's objectives, with disastrous results. In perhaps the most notable example, Xerox Corp. failed to capitalize on the computer mouse, the graphical interface and other major elements of computing that were invented at its Palo Alto Research Center.
Those freewheeling days are fading fast, largely because competition in technology has intensified. Labs must generate more breakthroughs that will quickly lead to new products in their companies' core lines of business.
Hossein Eslambolchi said that when he took over AT&T Labs in 2001, 80 percent of its efforts were long-term projects with goals a decade off. Only 20 percent was "direct research" with payoffs 12 to 18 months out. Now, he said, the ratios are reversed.
Shortly after taking over Hewlett-Packard Co. in 1999, Carly Fiorina ordered researchers to get more aggressive about filing patents, with bonuses of nearly $2,000 for employees who win them. Fiorina sought to broaden HP's royalties portfolio and support its new marketing slogan: "invent."
IBM boasts that it has gone even further in retooling its labs, as part of the company's shift into technology consulting and services. IBM researchers regularly meet with customers and IBM business units and even do some consulting.
IBM's senior vice president for research, Paul Horn, said in an interview in July that by comparison, rival labs were "building a bankrupt model."
But Horn concedes the pendulum might have swung too far. Perhaps with too many restrictions and time commitments, it's hard for IBM's geniuses to feel free and daring.
"We wanted to make sure that as part of the cultural change, we weren't becoming too focused on short-term results, and we still maintained an atmosphere of innovation," Horn said after Innovation Days.
Although Innovation Days was new to IBM, parts of it were not so innovative. IBM and other companies frequently bring in outside speakers to inspire technology ideas.
Last winter, Microsoft staged an internal "TechFest" in which researchers showed off projects to one another. In the spring, HP held its first "TechCon," inviting researchers and developers to a Colorado resort to discuss their work and hear motivational talks.
Intel Corp.'s chief technology officer, Pat Gelsinger, said the chip giant goes "out of our way" to expose its researchers to academics.
"When you get the process technology guy listening to the robotics guy, they get very bizarre, interesting conversations, but they find cross-disciplinary relationships that create new opportunity," Gelsinger said.
Innovation Days grew out of a pilot session last spring at IBM's Almaden Research Center in San Jose, Calif.
After that self-examination, Almaden scientists arranged a special fund for long-term projects that fail to win immediate support from the bean counters who dole out IBM's $5 billion budget for research, development and engineering.
They also called for meeting-free days so they could have fewer distractions, and open office hours, like college professors, so colleagues would bounce ideas off each other.
During Innovation Days, some people suggested days without e-mail. That would encourage more of the direct collaboration of years past, when researchers would walk down the hallway and sketch out ideas on a colleague's white board, said Samer Takriti, a senior manager in mathematics.
In Haifa, Israel, IBM researchers suggested that everyone be encouraged to occasionally break lunchtime routines and sit randomly in the cafeteria, so people from different departments would eat together and discuss their projects, said Gal Ashour, technical assistant to IBM's Haifa lab director.
Pressed for time, many researchers groaned when IBM first announced Innovation Days. And some lectures were sparsely attended.
Still, many participants said they were pleased. Jim Wynne, an IBM laser researcher who codeveloped the process used in LASIK eye surgery, said one lecture inspired him to dust off some work he did a while back. Now he hopes to patent it.
Wynne said Innovation Days "has a chance to change the culture." But he added: "Whether the concepts that come out of this get put into practice - or it's just a one-week refresher course on inventing - remains to be seen."
Associated Press Matthew Fordahl in San Jose, Calif., and Associated Press Bruce Meyerson in New York contributed to this report.
On the Net: