Stanford is the poster child for porous boundaries between universities, startups and bigger companies. Perhaps it's the constant sunshine. Students always have a startup on the go, professors are serial entrepreneurs, alumni still attend seminars. The long running public CS547 human computer interaction seminar is a melting pot of silicon valley ideas and luminaries mixing with current and past students. When I returned to London at the end of 1999 it felt so much colder, both literally and metaphorically. Some of the world's top ranked universities and design colleges barely mixed with the nascent startup scene. Why was this? And could it be improved? It was one of the questions that led to the founding of the Catapults, and one of my first jobs on joining Digital Catapult.

Hermann Hauser has an incredible track record as an entrepreneur and investor, helping found Acorn Computers, ARM (now used in 100B chips), the Amadeus venture capital fund and many others. It was his report for the UK government in 2010 that suggested setting up the Catapult centres to address the "valley of death": the gap between early stage publicly funded basic research and privately funded research at the latter end of commercialisation. You may wonder at the wisdom of governments investing in research and innovation at all. Why can't free markets and companies take care of it? The answer lies in The Entrepreneurial State, Mariana Mazzucato's 2013 book. She shows how the private sector only tend to invest once the state has made the high risk bets. Her famous example looks at all the iPhone components initially funded by governments: microprocessors, GPS, multi-touch screens, Siri, LCDs, lithium-ion batteries, the Internet, the Web, and so on.

In the manufacturing or bioscience industries, it is possible to use patents to capture the value of intellectual property created by a publicly funded university research group. Larger industrial players can license the patents and invest further to bridge the "valley of death". We didn't think this model made sense for the networked world of software. In the technology sector, the value is typically with the people and the team. The invention is commercialised through a startup or spinout. It isn't just an idea moving into the commercial world, it is the inventors themselves. This means a constant flow of people between universities and companies should be encouraged and facilitated; porous boundaries as you see in the Bay Area. Many of the founders of today's technology giants left university courses early to start their companies (Larry and Sergey dropped out of the Stanford PhD programme to start Google, Mark Zuckerberg and Bill Gates both dropped out of Harvard to start Facebook and Microsoft respectively, Steve Jobs dropped out of Reed College to start Apple).

That's why at Digital Catapult my initial university engagement strategy focussed on encouraging the movement of people. We gave top researchers the opportunity to explore commercialisation and spend time with companies by doing projects as Researchers in Residence with us, funded by the UK research council's Digital Economy theme. We worked across Centres for Doctoral Training, helping PhD students understand the opportunities available in startups, mentoring them and providing internships. We campaigned (generally unsuccessfully!) for universities to demand less IP or equity ownership in student or faculty startups, to encourage more to start. We ran pitch days for university spin-outs from all across the UK, directly leading to new funding rounds.

Over time it led to relationships with around 40 universities, including especially close ties with the 5G Innovation Centre at the University of Surrey, the Cloud Computing for Big Data, Digital Institute and Open Lab groups at Newcastle University, the Alan Turing Institute, the Digital Creativity Labs at the University of York, the AI Centre at UCL, the Horizon centre at the University of Nottinhgham, InnovationRCA at the Royal College of Art and the Digital Economy theme at EPSRC.