Lab Results May Vary

Bell Labs has the history, and Google--where engineers devote 20% of their time to personal projects--has the buzz. But other models of corporate innovation are also showing results.


The Program: IBM Research's Exploratory Research Program

How It Works: Any of its 3,000 researchers can pitch ideas to IBM Research's management team. Ten teams per year, composed of three to five researchers, get the funding and time--up to three to five years--to realize their concept.

Payoff: What began in 2001 as a hunch by two researchers who wanted to break down language barriers is now Mastor, a real-time, voice-activated foreign-language translation software for PDAs and laptops. Last year, 10,000 of the units were deployed to U.S. military in Iraq for English-Arabic translation.


The Program: HP Labs

How It Works: Startup teams of 10 to 20 engineers pitch "big bets" to a management board. About 30 get green-lighted annually, with a hefty dose of financing. The engineers spend 100% of their time on their projects and pair up with the corresponding business units to bring the product to market.

Payoff: Too early to say. The nascent program designed by the unit's new director, Prith Banerjee, an academic and entrepreneur, just kicked off in December. The focus on "big bets" is a drastic shift for the lab, which formerly focused on 150 smaller-scale projects per year--think incremental printer enhancement instead of a next-gen machine.


The Program: Xerox Innovation Group's Technology Incubation Network

How It Works: The company's 600 engineers, scientists, and researchers can pitch pet projects to a rotating council of eight peers--not management--who determine the recipients of small-scale seed funding. If initial efforts are successful and there's a strong business case, the venture is incorporated into Xerox's RD&E budget for the following year.

Payoff: The 20-month-old program has funded roughly two dozen projects, ranging from a research lab on virtual world Second Life, where Xerox researchers can meet as avatars and share concepts, to working with the Library of Congress to digitize 1 million images in a new, more agile compressed format called JPEG 2000.


The Program: Corning Innovation Lab's Technology Council

How It Works: Every four to six weeks, its 600 research scientists can pitch an idea related to Corning's core markets--glass and ceramics--to a council of company scientists and executives, including the CEO. Proposals that receive funding are managed closely through early stages of development; a fraction with the best commercial prospects become formal full-time projects.

Payoff: In August, Corning unveiled ClearCurve optical fiber, which is far more flexible than typical single-mode fiber. The innovation--named one of Time magazine's "Best Inventions of the Year" for 2007--began back in 2003 with two research scientists studying how to remove air bubbles from glass.

Add New Comment