Contrary to pop lore, Thomas Edison didn't invent the first lightbulb. That was Sir Joseph Wilson Swan, a British chemist, physicist, and inventor. Edison only came up with one that could fire for 600 hours instead of forty, making it the commencement commercially feasible "electric lamp" in history.

For scientists at General Electric's global research facility in Niskayuna, New York, that'southward more than trivia. It's a guiding principle for Anil Duggal, 38, an affable chemical science researcher who is trying to pull off an Edison-like feat. Duggal and a cantankerous-disciplinary team of scientists at the center want to develop a new kind of electrical lamp using an emerging technology called organic calorie-free-emitting diodes (OLEDs), well-nigh easily thought of as light-up plastic.

Why? Call it creative destruction. Or, for the slightly more than jaded perspective, phone call it a Hail Mary pass to save the iconic but struggling GE Lighting business (now function of GE'due south $fourteen billion consumer and industrial division). In 2002, it lost all of its Habitation Depot business to rival Philips. That unmarried defeat wiped out a total 7% of the unit'due south annual sales. Business hasn't improved much since. And in a article line such as lighting, Duggal'south work likewise fits in nicely with CEO Jeffrey Immelt'due south push to foster innovations that let GE widen its margins with difficult-to-re-create products rather than competing on incremental improvements and cost.

What's well-nigh striking about GE's renewed interest in innovation is non and so much the technologies themselves, but how the company gets them from lab to marketplace. "Equally a scientist, y'all take to figure out what makes this place tick. And it's not just applied science," Duggal says of GE'southward research eye. "If you can't sell a project, so y'all're going to accept a hard fourth dimension here." So OLEDs may not exist GE's most cut-edge research project — that might be its nanotechnology or molecular-medicine efforts. Simply it is a engineering that may ane day save GE's flagging lighting business organisation past, ironically, driving a stake through the lightbulb equally we know it. So OLEDs provide an intriguing window onto how the visitor integrates long-range research into today'southward strategic planning and how new ideas become through the system without getting thwarted, blocked, or worse. How do y'all usher in a game-changing innovation that'south years abroad from completion? Here's GE's shot at solving one of the oldest issues in business.

OLEDs are organic versions of LEDs — the stuff that makes our cell-telephone buttons light up — with a couple of key differences. For starters, instead of emitting a single bright point of light like LEDs practice, OLEDs produce a patch of low-cal over a wider area. 2d, OLEDs consist of a thin, flexible, plasticlike material, dissimilar LEDs, which are fabricated as rigid semiconductor chips — a process that would be laughably expensive for general lighting. Duggal hopes that the sparse textile will ane mean solar day be printed using a cheap roll-to-roll procedure, as newspapers are. That flexibility opens up a range of imaginative possibilities, from lit ceiling tiles that would supplant fluorescent overheads to illuminated curtains.

Duggal began to take serious notice of OLEDs nigh v years ago. As a chemist in corporate R&D, he had a general interest in the technology as far dorsum every bit 1989, when it was pioneered at Cambridge University. Information technology wasn't until scientists showed that OLEDs could produce a white light, spelling opportunity for general lighting, that Duggal's interest was fully piqued.

"Every year, OLED performance was getting exponentially better," he says. "One day I started plotting lines on a piece of paper showing this exponential growth, comparing it with conventional incandescent and fluorescent lighting technology. And you could see the lines commencement to intersect. So I'm thinking, If nosotros're a lighting company, there'due south both an opportunity and a threat here. This is a engineering that could overtake u.s.a., and all of a sudden our own lighting business organization becomes meaningless."

Inspired (and frightened) by his sketch, Duggal began chasing some seed money within the company to see how OLEDs might fit in with GE's existing businesses. He was able to convince a handful of executive managers at the lab, and he got the coin he needed to start. But to boost the project's initial funding and its staff from but Duggal to a squad of five, he enlisted the help of Greg Chambers. At the time, Chambers was one of GE's business program managers — a liaison between the global enquiry center and one of GE's business concern units. Together, they secured a grant from the U.S. Section of Energy to study OLEDs as an efficient lighting culling.

When Immelt took the reins in 2001 with a renewed focus on innovation, Duggal knew that this leadership alter was his opportunity. Under Jack Welch, who considered investing in new technology to be a "wild swing," OLEDs would accept likely remained a side project. Now they could exist a full-time calling. Information technology became a matter of Duggal building support for the technology amidst scientists and engineers besides every bit the marketers and sales forces in the business units. "They said, 'Yes, we recollect it's important for GE to exist in this,' " says Duggal.

When it came time to nowadays OLEDs as a candidate for advanced technology funding, Chambers quips that he went in and told his bosses that "GE should invent a new lightbulb every hundred years or so."

Turns out, he was only half joking. While the OLED market for displays in products such equally prison cell phones is a crowded and competitive i, with at least 25 serious players chasing it, nobody but Philips and Sylvania seemed to exist going after the full general lighting market. "A lot of low-hanging fruit is nevertheless available in lighting," Duggal told people at GE at the time, referring to the relative lack of competition. The projection won blessing equally an advanced technology program, one of 6, and as a result, its funding increased into the millions, and the staff grew to xxx. "The places that we've decided we can accept a pretty long-term look and take on a fair chip of risk are in what nosotros call 'sustainable industries,'" says Scott Donnelly, senior VP for corporate R&D. "I have a very hard time thinking that people aren't going to need energy 20 years from at present."

OLEDs still face a long road. "Yous may start to see OLEDs showing up in niche architectural lighting in viii to ten years," says Kimberly Allen, the manager of applied science and strategic research for iSuppli/Stanford Resources, a market enquiry firm. "Just you lot won't run into something similar flexible lighting for 15 or even 20 years." In fact, Allen says, the size of the OLED market for general lighting is incommunicable to forecast considering and then many technical questions remain unanswered.

This isn't news to GE. Proving commercial viability becomes more of import as OLEDs develop. Scientists do their part through a procedure called "toll gating," in which they incrementally tackle risks that a promising applied science will turn out to be a dud. Right now, Duggal's team is at piece of work on improving the brightness and elapsing of OLED lighting. In March 2004, they successfully demonstrated that their image could lucifer a standard incandescent bulb's light output. But the ultimate goal is to produce an OLED about seven times better than that, superior to even fluorescents.

The other major hurdle is solving the production problems that would let y'all print OLEDs as you would a paper. OLEDs demand to be hermetically sealed to work. Today that'south achieved past sandwiching the OLED material between two layers of glass. That's why information technology's premature to talk virtually roll-to-roll production. "In a style, those lines I plotted that showed the potential of OLEDs were deceptive, because there were things missing — similar manufacturing," says Duggal. "These are challenging issues. But we can either exist scared by them or find a style to work around them."

While Duggal's team works on those issues, a business concern unit's sales and marketing teams gather input about the technology from likely future customers. The net issue of this dance is that a technology must continually prove its marketplace mettle if information technology hopes to go on forth GE's path from lab to market place. In theory, that means no new GE applied science will hitting the marketplace unless there are customers already lined up.

How then does a technology like OLEDs non go lost in the strategic shuffle at a behemoth like GE, with $134 billion in annual sales? The respond lies in the visitor's yearly strategic planning process known every bit the S-i.

Contained within the consumer and industrial segmentation's Due south-one is a category labeled "innovation," under which projects fall into one of two subcategories: big bets and breakthroughs. OLEDs are listed equally a breakthrough now, meaning they're still a corporate research project. In one case technically viable, OLEDs would probable move to a big bet, where funding and technical development get handed over to the business unit of measurement. "If consumer and industrial were to go to corporate with their strategic plan and non have OLEDs on there, Scott Donnelly and Jeff Immelt would ask, 'Why are nosotros investing in this engineering science if there doesn't seem to be a home for it in the time to come?'" says Todd Graves, the business concern program manager who is at present responsible for shepherding OLEDs.

The concluding stage of the development process for a applied science similar OLEDs comes when consumer and industrial'south innovation squad — a group created four years ago to smooth a projection'southward transition from corporate inquiry to the business unit — hands it off to the new-product division. The scientists have eliminated all the invention risks. Now another team of engineers transforms the raw technology into a marketable product, a process that typically takes 12 to 18 months, according to Kevin Nolan, consumer and industrial's general managing director for new-production introductions.

Duggal realizes the challenge of ever getting that far. "It's very possible that OLEDs won't piece of work," Duggal says coolly. "Of course I'd be disappointed. Just if I wanted something like shooting fish in a barrel, I wouldn't accept taken them on. As a young scientist who goes into manufacture, instead of a university, you want to change the world. I think OLEDs are something that can change the world." Edison couldn't take said it improve.

Fast Take: GE's Rules for Innovation

Think Like Edison.

Thomas Edison was an inventor out to solve problems. Don't set out to invent new technology merely for applied science's sake. Instead, define innovation in terms of a riddle y'all are seeking to solve.

Big ideas happen at the fringes.

Recruit and retain talent from a broad spectrum of technical capabilities. When big brains expect at problems in a unlike manner — for example, physicists taking on chemical science bug — that's when breakthroughs can happen.

Make innovation pay its fashion.

Business units should pony upwardly for enquiry they actually want rather than headquarters doling out all of R&D'due south upkeep, hoping for a breakthrough. Requite early funding only to a scattering of promising technologies.

Set intermediate goals.

When managing long-term innovation areas, periodically review technical milestones along the manner to monitor progress. Prioritize the importance of each technical problem to solve, then chip away ane at a time.

Bet on the industry, not the technology.

GE tin can't predict amend than anyone else near when, or even if, game-changing technologies volition take off. But by focusing on an industry such as free energy, it can identify big bets within it in an effort to stay in the atomic number 82.

Ryan Underwood is a Fast Company staff author.