In the News
JULY/AUGUST 2006 (Vol. 23, No. 4) pp. 93-97
0740-7459/06/$31.00 © 2006 IEEE

Published by the IEEE Computer Society
In the News
Greg Goth
Global Technology Transfer Infrastructure Is Maturing
In 1994, the US National Science Foundation awarded Stanford University researchers US$4.5 million to explore new methods of accessing material on the World Wide Web. One project originating from that grant was the brainchild of a Stanford graduate student who was devising a slightly different way of giving priority to Web search results.
That new approach to Web search became the foundation for Google. Slightly before the close of the first quarter of 2006, Google's market capitalization value was $100 billion—meaning Google's shareholders benefited from the original $4.5 million federal investment at a ratio of 22,000 to 1. Less tangibly measured benefits, such as how Google's technology lets people work more quickly and efficiently from anywhere in the world with an Internet connection, go far beyond the cut-and-dried bottom line.
Of course, Google might be the ultimate example of government-funded research that hits the jackpot. However, even smaller advancements in government-funded research can yield significant dividends for the researchers' institutions. These dividends can also extend to the companies that partner with the researchers or that might be spun off from such research, to those companies' employees, and to those who take advantage of end products based on that research.
For example, while Google might be the most famous example of Stanford-incubated technology, the university has benefited greatly from the licensing infrastructure available to US-based researchers. According to Jon Sandelin, senior associate emeritus of the Stanford Office of Technology Licensing, the university has earned an ever-growing income stream from royalties derived from its researchers' efforts since the office was established in 1969. From 1969 to 1980, the total royalty income was $4 million. From 1981 to 1990, it was $40 million. From 1991 to 2000, it exploded to $400 million.
"However, the important point is that a great majority of the $400 million can be traced to inventions disclosed in the 1970s," Sandelin says.
The success of the US system of funded technology transfer hasn't been lost on the rest of the world's research community. From Stockholm to Shanghai, researchers, scholars, and policymakers are crafting legislation allowing for equitable distribution of intellectual-property rights arising from funded research. In addition, they're helping to build long-term national and multinational frameworks in which the public, nonprofit, and private sectors can each reap public investment's benefits.
Growing pains
However, much of the worldwide cultural and political discussion surrounding the benefits and challenges of technology transfer doesn't hinge on empirical evidence of sufficient returns on investment. Instead, it relies on an approach that demands short-term returns while instilling fear that somebody else is spending more and doing it better. This naive approach has two main causes. First, the legal infrastructure surrounding modern technology transfer is only 25 years old, and, as Sandelin's data shows, a given discovery's true payoff might come years after it's released. Second, the diverse political and cultural factors influencing each nation's and region's pattern of research funding makes statistically accurate direct comparison difficult, if not impossible.
For example, US researchers and policy analysts have expressed concern about the rapid increase in China's government-funded R&D. However, scholars familiar with many of the ground-level details say the Chinese system faces plenty of shortcomings. These shortcomings range from the difficulty of determining who truly owns a given discovery to that of determining which discoveries might actually prove profitable.
Meanwhile, European leaders have been using the US as the benchmark in their plans to raise what they see as laggard public R&D investment. European research initiatives don't seem to suffer from lack of legal infrastructure. In fact, analysts believe that one impediment to a more efficient European effort is the large number of cross-national and cross-framework policies that must be bridged to foster global European competitiveness.
US law is the wellspring
Worldwide, the consensus among technology transfer experts is that the 1980 passage of the US Bayh-Dole Act marked the beginning of technology transfer's modern era. In its 2003 final report on the status of technology transfer in the US (, the President's Council of Advisors on Science and Technology (PCAST) summarized the law's major provisions:

    • Nonprofit institutions, including universities, and small businesses may retain title to innovations developed under federally funded research programs.

    • Universities are encouraged to collaborate with commercial enterprises to promote the use of inventions arising from federal funding.

    • Universities must file patents on inventions they elect to own.

    • Universities must give licensing preference to small businesses.

    • The government retains a nonexclusive license to practice the patent throughout the world.

    • In cases of nonperformance, such as if a funding recipient fails to make sufficient progress in applying a "subject invention in field of use," the government retains march-in rights to license the technology to subsequent applicants.

Data compiled by the Association of University Technology Managers indicates dramatic increases in the diffusion of university-bred technology since the law was passed. On its Web site (, the association reports that
Before 1980, fewer than 250 patents were issued to U.S. universities each year and discoveries were seldom commercialized for the public's benefit. In contrast, in FY 2002, AUTM members reported that 4,673 new license agreements were signed. Between FY 1991 and FY 2002, annual invention disclosures increased nearly 250 percent (to 15,573), new patents filed increased more than 310 percent (to 7,741) and new licenses and options executed increased more than 365 percent (to 4,673). In FY 2002, AUTM members reported 569 new product introductions, and nearly 23 percent of their 26,086 active license agreements were associated with product sales by their licensees—dramatic confirmation that AUTM member licensees are successfully producing products covered by the license agreements described in this survey.
However, despite the impressive statistics that backers of publicly funded technology transfer can produce, some programs seem to be continually at the mercy of an apparently capricious and irrational political process. One example is the Advanced Technology Program (, administered by the US National Institute of Standards and Technology. ATP funds high-risk, high-reward premarket technologies on a cooperative basis with private industry and the university/nonprofit sector. For instance, even though the Research Triangle Institute's comprehensive analysis of the ATP Component-Based Software Development program documented an impressive payback of $250 million for a $42 million government investment (, the entire ATP program has faced elimination almost every year of its existence (and funding was once again cut in the initial budget request for fiscal year 2007). Although funding has been restored every year, ATP backers still need to fight behind appropriations committee doors, despite evidence of the program's positive cost/benefit ratios. This shows that even in open societies in which information flows relatively unfettered, decision making is often convoluted and confusing.
Asian climate: Committed but immature
China's rapid increase in funding high-tech R&D started with the announcement of the 863 Program in 1986. As the program (and internal political discussion around it) matured, the Chinese government formed the National Center for Science and Technology Evaluation in 1997, and NCSTE officials say they have been working on improving evaluation methodologies ever since. However, analysts familiar with the Chinese initiatives say the effort is greatly hampered by the limits of political debate in China as well as an immature foundation surrounding legal issues such as intellectual-property rights and conflict of interest.
In a 2005 study titled "Promoting IPR Policy and Enforcement in China," a working group commissioned by the multinational Organisation for Economic Co-operation and Development wrote,
The current system is the result of the two amendments to the Patent Law. The first, in 1993 granted PROs (public research organizations) the legal basis to own the IP generated by their research. The second in 2000 recognised the legal status of researchers to possess IP rights of their research through contracts with their PROs. The new ownership policy and related government measures, including the exemption of business tax and business income tax for technology transfer activities, encouraging the setup of technology transfer offices in universities and PROs, and the establishment of national technology centres, etc., resulted in increased technology transfer. However, due to a number of factors, including the weak awareness of IP and the lack of competence of IP management, the amount of intellectual property protected by universities and PROs remained small in both quantitative and qualitative terms, and the pace of institutional and capacity building fell behind the need for the commercialisation of public research.
As Zhang Jingan of China's Ministry of Science and Technology pointed out, further improving and reinforcing IPR policies will remain a focus of the government's efforts aimed at improving the innovation environment, establishing new innovation mechanisms and upgrading national innovation capability.
Stanford's Sandelin visited China as part of a delegation in 2004 and witnessed some of these issues firsthand.
"Whenever we'd ask if they had any kind of policy, they'd always kind of nod yes and say, 'Yes we have policies,' but they could never show them," he says. "There are no rules as to who owns what, and that makes it a challenge. … Another issue they face in China is that the research community is facing top-down pressure from the government to be productive, but there is no equivalent of our technology licensing offices. So they're accumulating large pools of patents, but it's questionable what value they really have."
Japan's legal and political infrastructure surrounding technology transfer is far more solid than China's, Sandelin says, but the country enacted the applicable laws only in the past decade ( So, just as it took Stanford 20 years to fully reap the benefits of royalty income after Bayh-Dole, Sandelin says it will also take time for Japan to realize such gains in the best of circumstances.
Europeans honing evaluation methods
In Europe, both national and multinational bodies are moving forward with new rounds of publicly funded precompetitive technology and with technology transfer activity. They're also establishing new structures to evaluate best practices and sufficient payoff on public investment.
In the UK, projects such as the Department of Trade and Industry's Knowledge Transfer Partnership (, which pairs private industry with the university and research sector, have yielded impressive results. According to the KTP, for projects in 2001 to 2002, participating companies that spent around £32,000 over two years made recurring annual profits of £156,000 before taxes (
The DTI also formed a Technology Strategy Board in October 2004 and charged it with evaluating the effectiveness of nationwide funding programs. In its first annual report ( technologystrategy/Technology_Strategy_ Board_Annual_Report_2005.pdf), the TSB noted that although the program's overall direction was satisfactory, "the project leads were generally unable to quantify expected returns on investment although their expectations of both the potential returns and the risks to achieving them both appeared to be high."
The TSB has therefore recommended that the program might better concentrate on larger projects. "This will require more focused competitions with appropriate resources earmarked for excellent projects," the board concluded.
On a multinational scale, the European research community has extended its ITEA (Information Technology for European Advancement) program to 2015. The second segment of projects, under the ITEA2 banner, will focus on research in embedded software systems. In a midterm assessment of the program, analysts from the firms IDATE and TNO recommended better coordination of communication between ITEA and the European Commission's Framework Program and more open and efficient national-funding decisions ( attachments/37/ITEA_MTA_Executive_Summary.pdf):
Funding decisions should be made more quickly and efforts should be made to synchronise national decisions as much as possible. Secondly, the policies and mechanisms in each country for managing the relationship between Eureka and national programmes should be made more transparent and available to all current and potential ITEA participants.
European analysts have also observed that the cultural differences between the continent's more comprehensive social-welfare structure and the US entrepreneurial spirit work against European interests. The US public and private sectors' greater acceptance of risk is an attitude these analysts believe must also be fostered in Europe.
For example, in a comprehensive report comparing the US and Swedish systems (, Swedish analyst Magnus Karlsson wrote that one of the greatest challenges to R&D-driven growth in that country resulted from a less-entrepreneurial business culture.
"Most analysts agree that the explanation for this situation includes the lack of venture capital, particularly from private sources (business angels), and the lack of management competences to start and grow businesses," Karlsson wrote. "The underlying factors that are pointed out are usually a less developed entrepreneurial culture and the lack of efficient incentives for investors and innovators. The primary setting for entrepreneurial activities in Sweden has traditionally been inside the large companies."
The European Commission has also funded CISTRANA (Coordination of Information Society Technologies Research and National Activities,, an initiative to facilitate improved multinational communication and exchange of best practices and evaluation data. In late 2005 and early 2006, the organization sponsored monthly workshops; one of its recent keynote speakers, the University of Manchester's Luke Georghiu, summed up the current state of technology transfer evaluation. He noted that to focus on merely the hard payback handicaps the champions of funding R&D. The failure to appreciate the full extent of both the private and the social returns from R&D is a key reason for the international decline in government investment in research, he claimed.
Measuring the qualitative payoffs
Georghiu's message hasn't been lost among those who agree and who possess data to back up that sentiment. AUTM recently began the Better World Project (, meant to communicate technology transfer's obvious and not-so-obvious worldwide impact.
AUTM president John Fraser says, "We initiated the Better World Project to broaden the look at the measures of success, and at the end of the day it comes down to this: this activity is successful if we can measure the number of lives saved, the quality of life impacted, and the productivity increases. The problem with this is, if you look at it as a business model, it's not very efficient. But there are a lot of things in the marketplace that don't generate positive cash flow but have major social impact."
As difficult as quantifying "soft" dividends might be, the research community knows the effort to do so isn't optional. How successfully such efforts as CISTRANA's and AUTM's convince skeptics that public payouts lead to payoffs unimagined might tell the tale of which societies thrive and which stagnate in the coming decades.
SourceForge Awards Highlight Open Source Trends
Greg Goth

People interested in grassroots software development trends have a new resource in the Community Choice Awards (
Nearly 250,000 people cast votes for the first awards, which were announced in April 2006 at LinuxWorld Expo. plans to make the awards an annual event.
"We just wanted to highlight the contributions open source software has made to the general good," says Jay Seirmarco, general manager.
The awards coincide with plans for enhancing the Web site and organization. Seirmarco says that notable among those enhancements will be advanced search capabilities to let site visitors find projects more efficiently, improved mail architecture to allow developer-to-developer email, improved internal-systems management, and more employees as the site's mission expands.
Seirmarco says the site's demographics have shifted markedly since its launch in 1999 as a repository for a core set of open source development tools. In analyzing the site's most recent statistics, he estimates that about half of the 24 million unique visitors per month don't fit the traditional perception of a developer.
"The blurring of the lines between the developer community and the user community is happening already," Seirmarco says.
Where users are developers
Two of the 14 award-winning projects embody that expansion and blurring as well as the potential for open source applications at all levels of the software stack to affect commercial markets. The winner of the best overall project prize, Azureus, is a Java BitTorrent client, with a dedicated community of users who also write plugins—more than 40 so far—to enhance its utility.
Azureus's developers inhabit one of the industry's most active new commercial segments—the infrastructure supporting large-scale file sharing. This segment includes both open source technologies such as Azureus that support peer-to-peer applications such as BitTorrent (which is also open source) and proprietary technologies meant to ease Internet service providers' administrative and bandwidth workloads. As this burgeoning part of the network architecture shakes itself out, Azureus CEO Gilles BianRosa says ISPs will have to work together to allow "some market-based solution to emerge."
BianRosa believes such a solution might be an open protocol for caching that will expose its API to various features, thereby encouraging a broad development landscape around the core technology. To capitalize on the new market opportunities, BianRosa says the Azureus team wants to make its technology more appealing to a wider audience by simplifying its user interface. They also want to take advantage of its current popularity by

    • better integrating features such as an RSS reader,

    • expanding its capabilities beyond its current status as a tool dependent on other services, and

    • forming a publishing community encompassing users from home videographers to professional filmmakers seeking distribution for their work.

The rise of scripting languages
A Web-based MySQL dashboard called phpMyAdmin is an example of the strong showing of scripting languages in the voting, in this case PHP. It won in both the database and systems administration categories. Its lead developer, Marc Delisle, is a network manager at Cégep de Sherbrooke, a college in suburban Montreal.
Although phpMyAdmin remains consistently in the SourceForge Top Ten and gets downloaded between 7,000 and 10,000 times a day (receiving between 400,000 and 600,000 hits daily), Delisle says there are no plans to build a software empire behind it.
"It's difficult to really want to cash in for such an application, because our team members already have a job, and we cannot afford to be forced to do something," he says.
He also states that the project's development team feels a sense of duty to its community, which ranges from Web forum hosts to database administrators.
"We have 200 feature requests waiting for us right now," he says. "We have a very small team, but feel somewhat responsible."
As to the new wave of applications based on scripting languages, Delisle says that while the relative ease of writing such applications might mean a larger pool of projects, that doesn't ensure an audience.
"There are lots of new applications, but their market quality varies," he says. "So it's kind of hard to remain popular for long. We've been successful for eight years, and for this, we are grateful to the community of users."
The anonymous developer
The contest's other two-category winner, Xbox Media Center, is a throwback to the days of developers using monikers to shield their identities. XBMC, which allows playback of a wide range of video, picture, and music formats on user-modified Microsoft Xboxes, took top prize in both the multimedia and game categories.
A member of the XBMC development team, "Pike," says the project embodies the community's reaction to proprietary formats, in which marketing considerations trump utility for end users.
"That's when independent developers or teams who don't have these financial or moral considerations at all, and just wish for something that works universally, happen to 'hack around' and make this happen," he says. "To me, they often end up with the best compromise. And then you have convergence because a coherent piece of software, or hardware, or both, is able to cope with most of the existing formats in a transparent way."
Pike also says that despite the commercial readiness of much open source development, the underground nature of XBMC will likely keep the project a pseudonymous hobbyists' retreat.
"Team XBMC always operated in kind of a gray area with XBMC since we use Microsoft's Xbox SDK [the software developer kit, which they need to compile code], which can't be obtained legally unless you're a licensed Xbox developer. That's why team members keep a low profile; it was never about making money."
Spreading the open word
SourceForge's Seirmarco hopes that the awards might be one factor users consider when choosing which software best fits their needs. He also hopes that the results have stimulated discussion of how best to keep advancing the open source message. Among the projects he would like to tackle is more formal outreach to other open source repositories and projects.