The Route From Research to Start-Up

Entrepreneurial Edge


HARTMUT NEVEN came from Germany to the University of Southern California a decade ago to lead research efforts in computerized face and object recognition. The aim was to create a tool for the military and law enforcement agencies to identify enemies and suspects as well as for airlines and other companies to check passengers and visitors.

In three years, Mr. Neven, a physicist with a Ph.D. from Ruhr University, succeeded in developing software that could analyze facial features, skin tones and the iris of the eye to achieve such recognition. Then, with encouragement from the technology transfer program at U.S.C., he formed a company named Eyematic Inc. It later became Neven Vision, with the aid of $200,000 in seed money from William Woodward, managing director of Anthem Venture Partners, a Santa Monica firm, and some of Mr. Neven s own capital.

Connections to the university did not end there. The Information Sciences Institute at the University of Southern California has continued research in computer recognition jointly with Ruhr University in Bochum, Germany, and with support from Neven Vision. They build on top of the basic technology, Mr. Neven said.

Such collaboration among university research, start-up companies and financial backers is a long-term trend dating to the Bayh-Dole Act of 1980, which mandated that universities collaborate with commercial concerns to promote use of the inventions arising from federal funding.

Many universities now have offices of technology transfer looking to turn research into commercial ventures at the same time that advances in information and biological science and the growth of Internet commerce has led to an explosion of entrepreneurial companies.

We see it as transferring back to the public benefits of technology that in many cases has been developed with public money, said Krisztina Holly, vice provost of U.S.C. who is leading a major expansion of its technology transfer program.

Defense Department research grants in the 1990s, for example, benefited the early facial recognition research behind Neven Vision. And those grants inspired continued development. Ultimately, the technology will allow you to point your camera phone at a movie poster or a restaurant and get an immediate review of the film or the fare on your cellphone, which will tap into databases, said Mr. Neven, who foresees one billion camera phones in use worldwide by 2010. That envisioned future may have been brought closer last year when Mr. Neven s company was acquired by Google for a price that financial experts estimate at more than $40 million.

Technology transfer clearly has paid off for Mr. Neven and his company s financial backers. But what do the universities get out of it? The benefits to U.S.C. for backing the research include potential licensing fees from patents on the developing technology and the possibility of contributions to the university from successful entrepreneurs like Mr. Neven.

Lawrence Gilbert, senior director of technology transfer for the California Institute of Technology, put it this way: Universities earn fees and royalties from licensing patents, but the real benefit comes from contributions to the endowment if the start-up company is successful.

Caltech has spawned 80 new companies in the last decade, including Impinj Inc., which makes critical semiconductor-antenna chips that power radio frequency identification, or RFID, tags. Such microscopic tags make possible continuous tracking of products that enter American ports or sit on the shelves of Wal-Mart Stores or in medical vials in hospitals. Offering information that goes far beyond the familiar bar code, RFID is increasingly required by the Department of Homeland Security and used by major retailers and consumer product companies.

Impinj, which is based in Seattle, originated in research done at Caltech by Carver Mead, a founder of modern electronics, and a student of his, Christopher Diorio, who is now professor of computer science at the University of Washington and chairman of Impinj.

To start the company six years ago, Caltech took a small equity stake, said William Colleran, president and chief executive of Impinj. It was not investing in specific technology but in potential. Initially, we were working with a communications chip, Mr. Colleran said, but in 2003 when the government began to mandate RFID, we saw that our technology could be the perfect product.

Impinj, which is private, has 150 employees and tens of millions in annual revenue, Mr. Colleran said. It also has $75 million in venture capital from nine major funds and looks to one day sell stock publicly, he said, which would provide a return for all those investors, including Caltech.

The equity investment by Caltech was in lieu of fees for acquiring a patent usually about $20,000. Caltech also defers license fees until a company gets on its feet. Our aim is to help the entrepreneurs get started; they need to retain all the cash they can, said Frederic Farina, assistant vice president of Caltech s technology transfer office. Thanks to the success of generations of its scientists, Caltech holds 1,641 patents and has 395 pending.

University policies differ on such matters. Publicly financed schools, like the University of California system, are not as liberal as private ones like U.S.C. and Caltech or as freewheeling as entrepreneurs often prefer. Public institutions collect the cost of acquiring patents upfront from faculty members who are forming companies. But our goal above all is to help make innovations and products available for the public good, said Jane Moores, interim director of technology transfer for the University of California, San Diego. The university, renowned for medical and biological technology, made 300 innovations public last year.

A company formed in 2003 by three researchers there illustrates how a university environment can invent new technology. Tinnitus Otosound Products began with the research of Dr. Erik Viirre, a medical doctor, who found that tinnitus, or ringing in the ears, occurs not in the ears at all, but in the brain. As a person loses hearing, the auditory cortex of the brain seeks to compensate by generating more nerve activity, which causes the ringing, just as a phantom limb can have pain after an amputation, Dr. Viirre said.

To combat the affliction, an often debilitating ailment for more than 30 million Americans, Dr. Viirre consulted with two university colleagues, Jaime A. Pineda and F. Richard Moore, who were pursuing other research on auditory problems. They theorized that generating external sound in a machine to exactly match the frequency of the sound in the patient s brain could cause the original ringing to diminish or cease.

They tested treatments on volunteer patients and came up with Customized Sound Therapy, which is able to mitigate tinnitus. The university acquired a patent in their names and that of the university. The company has operated on about $550,000 in grants, refining its technology and conducting trials for the Food and Drug Administration, so that the therapy can be classed as a medical treatment as opposed to only a mitigation device.

If Tinnitus Otosound succeeds, the University of California, San Diego will receive a royalty of 5 percent of the company s revenue, which will go toward financing more research.

This column about small-business trends in California and the West appears on the third Thursday of every month. E-mail: