The New Face of R&D at IBM, HP and Microsoft
Is R&D in the U.S. losing focus, or just shifting focus?
Pundits in recent years have taken to bemoaning a retreat by U.S. industry from basic research in science and technology. And indeed, it's easy to cite research labs whose glory days have come and gone -- Bell Laboratories comes to mind. But consider this: IBM, Microsoft and Hewlett-Packard together spend $17 billion annually on research and development.
That's right, $17 billion.
While many of those dollars are directed at product development, hundreds of millions are flowing into areas like computational biology, nanotechnology and advanced mathematics that may take years to bear fruit, if ever.
It's significant that each of these companies has undergone substantial changes in its research labs recently:
Then, earlier this year, Microsoft Research, which adds 50 computer science Ph.D.s to its ranks every year, announced that it would greatly expand its research campus in Beijing and open a new lab in Cambridge, Mass.
While the three companies have strikingly different research agendas, they have one important thing in common. All three are increasingly reaching outside of lab walls to collaborate in research with other companies, universities and customers. With that outreach comes a new openness that can speed the flow of ideas into the marketplace, according to Henry Chesbrough, executive director of the Center for Open Innovation at the University of California, Berkeley.
A Short History of R&D
"R&D is basically seeking out new knowledge, and the question is, where are the good ideas?" Chesbrough says. "After World War II, the good ideas were loaded up in a small number of large companies -- Bell Labs, IBM, Xerox PARC, GE and so on. These were islands of towering knowledge in a relative desert."
At the time, he says, universities generally disdained working with companies and instead relied on a federal government that was eager to fund research that might help win the Cold War.
Then the Berlin Wall fell, and much of the federal largess dried up, Chesbrough continues. With its antitrust actions, the government turned its attention to reining in technology giants like AT&T and IBM. Meanwhile, Silicon Valley was born, and so was the Internet.
"The result was product markets got more competitive, and those big companies couldn't sustain the long-term investments in research that they could in the earlier period," he says -- their money went instead to competing in their markets in the short term.
Into this research breach stepped smaller, newer technology companies, universities, companies in Europe and Asia, and in some cases, even customers. To Chesbrough, that's all good news. "Today, no one has locked up the really good ideas, and the R&D processes of large companies have to connect to these parties and make use of them," he says.
In fact, HP, IBM and Microsoft are all currently showing a strong move toward a favorite research concept of Chesbrough, "open innovation." As Chesbrough spells out in his book of the same title, open innovation calls for good ideas to come from both inside and outside the company. In turn, companies take the fruits of those ideas to market through internal as well as external paths.
What follows is an overview of how three of the biggest names in research are putting open innovation and other concepts into practice in a changing R&D landscape.
HP Labs: Five Big Bets
Hewlett-Packard Co. last year hired Prith Banerjee, the engineering dean at the University of Illinois at Chicago, as the new director of HP Labs, overseeing 600 researchers in seven labs around the world. The company followed up with an announcement the following March that HP Labs would shift focus from a large number of smaller projects to a few "big bet" projects in five major research areas -- information explosion, dynamic cloud services, content transformation, intelligent infrastructure and sustainability.
"These are the big research challenges that we think are most important to our customers in the next decade," Banerjee says. Individual projects include exascale computing, social computing, quantum computing and so-called green computing.
Explains Banerjee, "We had taken the approach of letting 1,000 flowers bloom and hoping a few would pan out. [But] we were working on a large number of projects without enough resources on each one. We'd have two or three people on a project, but now we'll have 20 to 30 large projects, each with 10 to 20 researchers working in teams." As a consequence of this change, he says, product divisions at HP will get research prototypes from HP Labs that are more fully developed, enabling products to be brought to market faster and at lower cost.
Some observers of the new strategy at HP Labs -- including competitor IBM -- suggested it was yet another retreat from long-term basic research in favor of short-term, product-oriented work.
But Banerjee insists the opposite is true. In the past, he says, less than 10% of HP Labs' budget went to exploratory, or "blue sky," research. Under the new plan, he says, one third of spending will be on "exploratory research," one-third on "applied research," and one-third on "advanced product development."
A major thrust of HP's new strategy is increased collaboration with other companies, universities and venture capitalists. Toward that end, the company recently set up the HP IdeaLab Web site, which offers would-be partners sneak previews of research prototypes.
And in May, HP released a global request for proposals to universities seeking collaboration in each of the five major research areas. Banerjee says he's seeking real partnerships in these areas, not ad hoc, one-off joint projects such as HP might have done in the past. "Never in the past have we gone to the outside world with so much detail about what we are working on," he says.
To be sure, behind all the love of open innovation is bottom-line business sense. Such a strategy not only allows HP to cull ideas from a wider pool, it also allows the company to mitigate risk and share research costs.
Clearly, Banerjee has revenue in mind when he says, "Our approach gives us a sharper focus in the areas that we believe will help our customers addresses the challenges they will face in the next decade. It also ... speeds the time it takes to quickly turn breakthrough research into real products and services."
Either way, the change at the top of HP Labs says a lot about where the company is headed, Chesbrough says. "It's interesting that a company the size of HP would bring in, not a career engineer or scientist from the company, but an academic," he notes. "Instead of the go-it-alone attitude, I see this as evidence of a much more collaborative, distributed process. And that's a very good thing."
IBM: "Collaboratories" and More
The buildings that house IBM's Thomas J. Watson and Almaden Research Centers are emblematic of an earlier time and an earlier attitude toward research. "The places are built high on a hill far away from everything else," Chesbrough observes. "They are almost physically designed to keep ideas from leaking out. If you look at the newer ones -- like Microsoft, Google or even Intel -- they are constructed to get ideas to flow in rather than to keep them out."
Buildings notwithstanding, research is a new ball game at IBM, Chesbrough says. "IBM is an old dog that has learned some wonderful new tricks and is having great success doing it."
While IBM remains strong in basic research in materials, semiconductors and the like, it has turned its R&D efforts more and more toward services and support technologies, he says. And a decision to support non-IBM products in its Global Services unit has "pushed IBM Research into the open standards domain -- Linux, Java, blade servers and other things," he says.
Shortly after John Kelly's ascendency to the top of IBM Research last summer, he announced that IBM would spend more than $100 million over three years on each of four "high-risk" basic research areas:
Cloud computing and Internet-scale data centers
Integrated systems and chip architecture
Managing business integrity through advanced math and computer science
Kelly promised to fund another 15 research topics at $30 million to $50 million each, and many more at lesser levels.
He said IBM would increase collaboration with universities, government agencies and other companies, citing the success of earlier alliances with universities in nanotechnology and semiconductors. He also mentioned speech technology as an area in which IBM would collaborate more.
Part of IBM's new game plan is the establishment of "collaboratories," mostly small, regional joint ventures with universities, foreign governments or commercial partners designed to tap into local skills, funding and sales channels in order to get new technology quickly into the marketplace. For example, in February, IBM announced it would form a nanotechnology collaboratory with Saudi Arabia to develop and market water desalination, solar energy and petrochemical applications.
Mark Dean, an IBM Fellow and a vice president at IBM Research, hails the payoff from distributed collaborative research, but he says it has its challenges. "When you have people everywhere, you need technology to make people feel they are together, like there is some camaraderie," he says. "You need a sense of family on a project. This is an area that has to be worked on."
IBM is adding "big bet" projects in exploratory research, Dean says. "For example, how does DNA interact with carbon nanotubes for self-assembly of circuits?" he says. "[There is] also spintronics and real-time analysis of large amounts of data from sensors all coming at you at one time. These are things nobody has done before."
IBM is also increasingly collaborating with customers, Dean says. For example, it is working with "a prominent candy company" to apply a prototype Web analysis tool, called Business Insights Workbench (BIW), to find hidden patterns and meanings in structured and -- most important -- unstructured data. "The company wants to do business in some emerging markets," Dean explains, "and [BIW] will look at trends and biases within a culture to predict whether a particular brand of chocolate will be bought." Another big plus: The client company will be able to analyze Web-based information in any language without translation, Dean explains.
Microsoft Research: The University Model
While companies such as HP rely on partnerships with universities to gain access to many basic technologies, Microsoft Corp. is trying to create its own university, Chesbrough says.
Indeed, Microsoft Research can reasonably claim to have already done so. It has been led from its beginning in 1991 by Richard Rashid, a former professor of computer science at Carnegie Mellon University. Rashid makes no secret of his operating model for Microsoft Research: "The work we do is not that different from what you'd find at Stanford or Berkeley or [Carnegie Mellon University] in the sense that it is publishable basic research that is peer-reviewed."
Rashid adheres to an outlook that sets Microsoft Research apart from many IT companies where every research project has a product line in mind: "Our research may have a short-term impact on the product groups, but that's not why we do the work -- it's a consequence of the work."
For example, Rashid says, he started a small research group in computer vision -- machines that can see -- in the mid-1990s when no Microsoft product at the time seemed to need that kind of technology.
"But in a few years, digital imaging and photography became a huge value to other parts of the company in areas like photo processing, image analysis and signal processing, and things like Windows Media and audiovisual codecs came out of that earlier work," he says. "And now things like Microsoft Surface are all based on computer vision technology. In fact, you could have read about the work leading up to Surface five years ago."
Because Rashid's philosophy is to first do good computer science, and then see where it might fit, he focuses first on people. "My biggest lever is who I hire and who I fire; it's not telling people what to do," he says.
Asked why he decided to launch a new lab in Cambridge, Mass., that will work at the intersection of traditional computer science and the social sciences, he says, "You don't establish a lab without the right person to do it. We had a great researcher, Jennifer Chayes, and she was really excited about a lab in that area [of the country]. She's fabulous and has done incredible work. If it wasn't for her energy and initiative, it probably wouldn't have happened."
To be sure, many of the 272 research projects named at Microsoft Research's Web site are structured with major product lines like Windows, Office or Xbox in mind. But many seem to have no likely application in anything the company sells today."We are growing outward into areas where computer science intersects with other disciplines, like AIDS research, computational biology, astronomy, earth sciences and the environment," Rashid says. "We are increasingly engaged where computer science is making a big difference in the way other sciences are done."
Asked about the trend toward outreach and collaboration, Rashid reels off a long list of global partnerships, many of them formed in the just the past two years: a joint lab with INRIA, the French national institute for research in computer science and control; a partnership with the University of Trent Center for Systems and Computational Biology in Italy; a partnership with the Barcelona Supercomputing Center in Spain that focuses on parallel distributed processing; the Microsoft-Intel Universal Parallel Computing Research Centers at the University of California at Berkeley and the University of Illinois at Urbana-Champaign; the Microsoft-sponsored Center for Computational Thinking at Carnegie Mellon; the partnership with Georgia Institute of technology at the Institute for Personal Robots in Education; the partnership with universities in the Latin American Collaborative Research Federation; and the annual Microsoft Research Faculty Summits.
Universities Shift Focus, Too
Those kinds of partnerships, at Microsoft Research and elsewhere, have been spurred in part by a shift in thinking on university campuses, where working with corporations was once viewed as somehow crass and unworthy. "The model you'll see going forward," Chesbrough says, "is for universities and their leaders to become much more skilled and comfortable working with companies, and it will be much deeper partnerships."
For example, HP is working with the Real-Time Computing Laboratory at the University of Michigan to develop algorithms for the dynamic allocation of resources -- such as CPU cycles, memory and bandwidth -- across applications in virtual environments in order to meet specified service level objectives.
Kang Shin, a computer science professor and the founding director of the lab, says the university is well suited to provide the theoretical and mathematical underpinnings for the algorithms, but it lacks the real-world experience and the huge testbed -- up to 50,000 servers in a data center -- that HP brings.
"We work side by side with HP, and we write papers together," Shin explains. "We work together so that we define real problems, not nonproblems. I think that's more important than the money we get from HP."
Despite all the advantages, there are some downsides to open innovation, Chesbrough acknowledges. "A software company from the early 1990s called Go Corp. lost its pen-based computing operating system intellectual property and market position by sharing too much with Microsoft," he recalls. "However, most companies I have encountered are, if anything, overly aware of these risks and use these risks to excuse themselves from opening up their innovation process."
To Chesbrough's mind, that's a mistake. "The many successful stories of companies adopting open innovation show that you can be more open without giving everything away."