Market opportunities offered by e-business could mean that research assessments will need to include a measurement of the success of technology transfer.
Brian O'Neill, from the electrical and electronic engineering department at Nottingham Trent University, told a Brighton conference last week that despite the financial hurdles involved in setting up spin-off companies, he felt strongly that the days ought to be numbered of engineering departments having their research assessed by the number of research papers published.
"Success in technology transfer is an important measure for the quality of research," Dr O'Neill said. "It can certainly prove a great advantage to students.
"The bottom line of anything I've done is the training I give to research students. They've all got jobs at a far higher standard than they would have if they'd just come out as raw graduates."
Dr O'Neill, at "The Role of Physicists in Building the Internet" conference, held as part of the Institute of Physics' annual congress, said it was possible to carry out small-scale research within a university that can compete with larger industrial research groups.
But researchers must find niche applications for their products and be prepared to be more adaptable than larger companies.
Dr O'Neill founded his spin-off company in 1996 after a research project involving digital circuit design led to a commercially viable microchip. The chip is used to link microprocessors, acting like the hub of a telephone network. Its main application has been in state-of-the-art computer graphics animation, but a new version of the chip could play a large part in creating the first "home networks", connecting household devices to the internet.