Restructuring of HP Labs changes global profile
March 24, 2008
By Joaquim P. Menezes, Toronto
"[HP Labs] is one of the few places where's there's still a lot of 'R' left in R&D."
HP's chairman and CEO Mark Hurd made this claim recently, during what he called the "re-launch of HP Labs"—an event held at Cupertino, California recently.
"Most of the R&D in the world today is D and we've got a commitment to the R," Hurd went on to say.
An interesting statement, considering that until recently, less than 10% of research projects at HP Labs were "Blue Sky" viz. research not targeted at short-term product development, but expected to deliver phenomenal "long term" benefits to the company…and society.
The 10% statistic was provided by Prith Banerjee, HP Labs director, who was appointed to this post last August.
But Banerjee, and other HP executives, promise a renewed focus on research at HP Labs—in fact this was a prominent theme in their presentations at the Cupertino event.
Here's a sampling of what they said on this issue:
Hurd: "With the re-launch of HP Labs we're placing fewer, bigger bets on the "R" that we think will have a bigger impact on HP Labs. This is a big deal for us." The HP CEO promised HP would be "a lot more transparent" about the "R" it is working on.
Banerjee: Moving forward "a third of our future research will be exploratory, a third applied research, and a third advanced product development."
But the executives steadfastly declined to provide even ballpark dollar amounts for projected expenditures in each of the three areas. Nor was there any suggestion that HP's total R&D budget would increase significantly, or that more funds would allocated for exploratory research.
So we have to make an educated guess.
We do know that the total number of R&D projects at HP Labs, will shrink significantly, across the board, from "hundreds" to "20 to 30 big bets."
Given that—it's more than likely that HP's actual expenditure on exploratory research will remain the same.
The moot issue then is: whether spending approximately the same amount on fewer "strategic" projects constitutes a deeper commitment to "R" on the part of HP.
It could well be argued that by focusing research time, resources and energies on fewer crucial projects, HP Lab researchers will come up with what author, speaker and venture capitalist Guy Kawasaki calls "curve jumping" innovations—which he defines as inventions that enable you to do things not just 15 or 20% better, but 10 or 15 times better.
Banerjee and other HP executives certainly expect this to happen.
Problem is the existing R&D environment in large technology companies isn't really conducive to such innovations—a fact noted by industry experts.
What we really witness today is many tech companies seeking swift and easily measurable returns on their R&D investments.
These expectations sometimes place the research director in a quandary.
On the one hand, there's pressure from the business side to deliver timely and tangible results.
But the use of criteria such as "discounted cash flow" and "ROI" to measure the value of research projects is also troubling to the research community, say analysts Graham Mitchell and William Hamilton in a paper titled "Managing R&D as a Strategic Option." "This is...because of a widespread sense that most investment analysis discriminates unreasonably against the longer term and more risky programs, many of which ultimately bring the largest benefits to the corporation."
They note that research directors, while appreciating the importance of longer term factors to their companies' growth and survival, are being forced to "argue their case with rules of evidence that often de-emphasize these factors."
Banerjee acknowledges this problem. "I am [aware] of the challenges facing corporate research in general—the struggle balancing long-term and short-term research."
But the new director of HP Labs says he resolves to strike the balance between longer-impact research and the quest for results.
"The one way we'll decide whether to fund projects is to determine if they can have a big impact," he says. "I want HP Labs to work on big-time, risky projects."
Are there any such projects in the works at HP labs? If so, what big-time benefits do they promise? These are questions I sought to get answers to at the HP Labs "re-launch" event in Cupertino.
Presentations by HP execs at the event and -- following these -- the demos in the "experience room" did highlight key HP Labs initiatives (some of which are still in their nascent phase).
These include:
- CloudPrint: a technology that allows you to store documents on the Web and retrieve and print them on any printer, using a mobile phone. This service—developed by Bernardo Huberman and Scott Golder at HP Labs—is free with no limit to the number of documents you can print. The beta version of the service is available in Canada, the U.S., and Europe.
- Foodsville: a community site created by HP and Applewood Books for food enthusiasts that allows them to read, contribute to, or modify, recipes from out-of-print cookbooks.
- Logoworks: Shane Robson HP's chief technology officer related how through Logoworks—the Web 2.0 graphic design service that HP acquired last year—HP got a logo designed for the 2012 London Olympics that's "better looking" than the official event logo, was delivered significantly faster, and was of an "order of magnitude" cheaper.
Interesting as all this was, following the massive build-up by Robson and Banerjee, what we actually got to see in the demo room was a disappointment.
There was nothing—for instance—to illustrate Robson's statements about the future of Search technologies, a future that HP Labs is presumably contributing to.
"There [will be] more intelligence in the Search, more understanding of who you are, where you are, what your preferences are—and this will give [you] the ability to target the information [you] care about," the CTO said.
This was a motif in Banerjee's presentation as well. "Our researchers," he said, "are working on technologies that will intelligently anticipate the services needs you have based on who you are, where you are, and what your preferences are." Wonderful concept. But I don't recall demos of any actual or upcoming HP products or services that illustrated this principle.
Even CloudPrint—indisputably a great technology—doesn't really exemplify "location awareness" and "needs anticipation"—both prominent themes in Robson's and Banerjee's presentations.
The same is true of the HP CTO's assertions about a seamless user experience becoming available across multiple devices and services.
"How do we ensure that for all devices we use we use, we have a consistent interface to the services that we care about," Robson asked.
How indeed?
Beyond the statement that this is happening or will happen, we didn't get any further insights into this during the event.
What I did find interesting, however, is the announcement that much of HP's R&D, moving forward, will be accomplished win partnership with outside groups using what the HP Labs director dubs the "open innovation" model.
HP, Banerjee said, would pursue relationships with venture capital firms, universities, startup companies, government agencies, and other partner companies to amplify its internal R&D.
While "open innovation" sounds like a fabulous idea, getting multiple organizations to work together on a research project is a very complex business.
As experts point out, many factors can interfere with this goal including:
- Issues around ensuring equity between partners in such relationships;
- The need to protect intellectual property and achieve a good return for time and money spent; and
- The fear that sharing ideas and resources, and opening up one's business affairs to another company might jeopardize competitive advantage.
This kind of collaboration is not something HP has attempted in the past, and given the inherent challenges, if the experiment succeeds for HP, it may well serve as a prototype for other large enterprises to adopt.
By Kripa Raman
Recently in Palo Alto, HP, the largest global manufacturer of printers and personal computers, has restructured HP Labs, its R&D division which has units in seven countries, including India.
The 23 HP Labs centers will now focus on 20 to 30 “big bang” projects rather than on the 150 being pursued currently, Prith Banerjee, director of the division, told reporters on a recent visit to Palo Alto sponsored by HP Labs.
Twenty to 30 researchers constituting “dynamic, high impact teams’, will work on each of the selected projects, he said.
Banerjee, who was appointed director of HP Labs eight months ago, comes from an entrepreneurial as well as academic background. He had founded two start-ups and also served as dean of the College of Engineering at the Univ. of Illinois in Chicago.
For the India unit of HP Labs, located in Bangalore, the mission is “investment for the next billion” he said. For example, this unit is working on a gesture-based keyboard where users can input data in their own language even though their keyboard is an English ‘qwerty’ one. This is a technology that HP Labs India had been working on even before Banerjee took over. Data entry is not made complicated because of the constraints of an English or even bilingual keyboard. This technology would be relevant to many countries that use a phonetic script.
Bar coding of certificates
A newer application from India on display at the conference at HP Labs’ Palo Alto premises was one that allowed bar coding of certificates, to prevent them from being tampered with or modified, something not uncommon in the country. Several universities and government offices have shown interest in this, said an HP researcher.
The display booth illustrated this, showing a certificate of graduation submitted by an applicant to a potential employer showing the applicant’s year of birth as 1965. But upon scanning the bar code on the certificate, the employer got a scanned picture of the original certificate which showed the applicant’s date of birth as 1960!
Another ‘print cast’ technology being developed in India allows people to use commercial cable and wireless networks to deliver educational content and print media for developing economies. This can be used for low cost, distance training, said Banerjee.
Aims at better returns
For HP Labs as a whole, the new blueprint for corporate research is aimed at better return on investment in research projects, with the right bets being determined with the help of an internal review board. But this is not to say that there will be no breakthrough research, said Banerjee.
One-third of the work will consist of breakthrough exploratory basic research, against 10% earlier. One-third will consist of application research and one-third advanced development research.
The 600 researchers at HP Labs will work on five major themes. These will be “Information Explosion” (how to find the ‘needle’ in the mammoth amounts of information being generated); “Dynamic Cloud Service” (providing reliable information from anywhere); “Content Transformation” (transfer of physical and analogue data into digital content and enabling of the same in the reverse direction too); “Intelligent Infrastructure” (smart devices networking that will operate at Internet speed and scale) and “Sustainability” (technology to reduce power consumption and carbon footprint on the planet).
HP Labs will also collaborate with venture capitalists, universities and entrepreneurs to exchange experiences and ideas. A website, HP Idea Labs, will allow researchers and partners to take a peek at some of the research happening too and help them develop on it.
For more information, visit HP Labs at www.hp.com/go/NewHPLabs
SOURCES: HP, Hindubusinessonline.com, Reseller News
|
