The Ecosystem: patent licensing makes way for partnerships in university tech transfer

Patents remain an important way for research-intensive universities to exploit their work, but the commercialisation scene is changing. Direct licensing to large companies appears to be in decline, with universities increasingly expected to develop technologies to higher levels of maturity and lower the risk, and new actors stepping in to support university ecosystems.

Meanwhile, in place of licensing deals, technology transfer increasingly involves partnerships, as companies seek to manage the risks associated with early-stage innovations in rapidly moving, multidisciplinary fields.

“One of the ways technology transfer is evolving is that we're seeing less interaction from large companies around licensing,” said Simon Hepworth, director of Enterprise at Imperial College London and the current president of TenU, an international collaboration between universities that shares best practice in research commercialisation. “Some of our counterparts in the US say they haven't signed a licence with a large corporate for the last five years.”

Where licensing is happening, it appears to be more with spin-outs and small and medium-sized companies. “For example, they may be companies that started at a university and, ten years on, they are looking to licence complementary technologies,” Hepworth said. “This might be because the corporates would rather avoid the risks involved in the early stages of technology development, but we need to ask more questions before we know for sure.”

Imperial is carrying out a licensing review in order to understand if something similar to the US situation is happening in the UK, and the factors that are driving these decisions.

The same trend is also apparent on mainland Europe. “A decline in patent licensing to large corporations is something we have also seen, for quite some years, but patent licensing in general is not declining,” said Paul Van Dun, general manager of the technology transfer office at KU Leuven in Belgium. “In Leuven, for example, we see more patent licensing to small biotech companies and to special purpose vehicles.”

These include venture builders and venture capital companies which take an active role in forming start-ups. And their appetite for risk appears to be greater than more established companies.

“Some of these smaller companies and special purpose vehicles are more willing to enter into early-stage development pathways,” Van Dun said. However, they also want something more from the university.

“When you license to a large corporation, they usually take over the technology development process to a large extent,” Van Dun said. “The smaller and medium-sized companies, and the special purpose vehicles, are looking for a more collaborative effort, and we conclude more license and collaboration agreements with them, rather than simple license agreements.”

 

Partnering

One driver is the increasingly interdisciplinary nature of technologies, for example covering both information technology and the life sciences. This can also affect the intellectual property (IP) aspects of the deal. “It’s not just a patent that is being licensed, but also things like non-patented cell lines or non-patented software,” said Van Dun. “So, you can see a slight trend towards package deals rather than simple patent licensing.”

The picture is slightly different for Silvio Bonaccio, the head of ETH transfer and ETH Zurich. He sees no decline in patent licensing, with interest from large companies holding steady. But he has seen increasing interest in partnering with academia, rather than simply buying a patent off the shelf.

“From industry, I see a trend for more collaboration with top-class universities,” he said. “That’s probably also because research has become ever-more expensive, and areas such as artificial intelligence are moving so quickly. Then it’s good for industry, and also maybe universities, to team up and work together.”

In these cases, the licensing is often an inherent part of the partnership agreement, rather than a separate transaction.

Among the many factors driving this trend is the cost of research. “Fundamental research is becoming ever more expensive, particularly if you need a lot of hardware,” Bonaccio said. “So, it’s about joining both the financial means that you have, but also the knowhow on each side, to produce something together.”

A recent example is the Boston Dynamics AI Institute, a US private research foundation, which is setting up an office in Zurich this year. It will be led by Marco Hutter, professor of robotics at ETH Zurich and director of its robotics centre.

“The AI Institute is interested in collaborating with us because there is knowhow on both sides that will allow this type of technology to be brought forward faster than would be possible by either of us working alone,” Bonaccio said.

ETH Zurich is increasing its focus on industrial collaborations, setting up a partnership for innovation group in 2023. With a staff of 28, this group covers everything from showcasing the university’s capabilities and managing initial contacts, to partnership agreements. “We are moving in that direction to see if we can get more technologies and knowhow from ETH into the real world,” Bonaccio said.

 

Spin-outs

Universities are also seeing an increase in spin-out activity, which in some cases can play a role in de-risking technologies. “Large companies may then be interested in becoming investors in the spin-out, to keep an eye on how those technologies develop, with the option of acquiring the spin-out later on,” Hepworth said.

Bonaccio agrees. “Spin-out companies are such a good vehicle for bringing early technologies to the market, or at least to the attention of others who can bring them to the market, such as big corporates who may be willing to buy them,” he said.

Universities are putting even more effort into supporting spin-outs and building internal teams that can take cover both the research and the business roles involved. “We are seeing more postdocs and more PhD students stepping in to become the business leaders of those spin-out companies,” said Hepworth. “The academics then stay in their faculty positions, continue their research and teaching, and maybe hold a part-time role within the company.”

This has led to schemes designed to support early-career researchers with entrepreneurial ambitions, such as Medtech SuperConnector and the AI SuperConnector, which Imperial runs with other academic partners.

ETH Zurich has also seen a significant increase in spin-out activity, with a record 43 new companies receiving its ETH spin-off label in 2023 for exploiting research carried out at the university. Artificial intelligence and biotechnology were both strongly represented.

Bonaccio attributes this to the longstanding support given to younger researchers who want to start their own companies. “This is now paying off, because people can see the success of others and want to do something similar,” he said.

ETH is planning to expand entrepreneurial support to take in the whole start-up community at the university, not just the ETH spin-offs based on university IP. A new group, ETH entrepreneurship, was recently set up to this end. “There are many more companies being founded out of ETH by students, and we want to create a bigger community,” said Bonaccio.

Universities are also developing technologies to a point where investors are willing to take the risk of backing them. “We have programmes like DT-Prime that provide active management of technology development, and much needed funding,” Hepworth said. “This involves sitting down with the principal investigator on a research project and asking them what is the most important experiment you need to run to demonstrate to your target sector that your technology is going to be commercially viable?”

For example, a catalyst might work perfectly on day one in the lab, but the real industry need is that the catalyst performs to within say 90% of its capacity after many months of use.

“So, it’s about creating real-world usage profiles that we can test our technologies against,” Hepworth said. ”We’ve also seen through DT-Prime that this exposure to specific industry needs provides really helpful insights to the academics, not just for their spin-out ambitions, but also for their future research agendas.”

 

Copyright

At the same time, the type of intellectual property is changing. “Copyright is becoming more important for us, simply because software (eg for AI) is becoming more important, and in Europe that is primarily protected by copyright,” said Van Dun.

It is also increasingly common to find software linked to more tangible innovations covered by patents, such as in robotic systems driven by AI. Combining several kinds of IP in a licensing deal, often across different disciplines, adds complexity for the technology transfer office.

“You need people in your office with expertise in engineering, software, etc, and with the ability to deep dive into the life sciences, which is more patent-driven,” said Van Dun. “And the more interdisciplinary projects you have, the more important it is that people with this different expertise collaborate closely together internally.”

Efforts are underway to establish best practice when licensing software, with TenU currently working on a guide to deal terms. “This will look at the special characteristics of knowledge transfer to software spin-outs and other companies, where the intellectual property is typically not patents, but code, knowhow and sometimes datasets,” Hepworth said.

The core business value in these cases can also be more about having a continuous interaction between the founders and their customers, rather than the foundational technology that came from the university.

Speed is often critical, particularly when it comes to AI spin-outs, and establishing best practice in licensing terms will help make commercialisation more efficient.

“At Imperial we've had a number of spin-outs, such as Contex AI, that formed and were sold within a two-year period,” Hepworth said. This can mean a radical shift in perspective.

“If you're in AI, the spin-out needs to be thinking about its exit at a detailed level almost at the point that it's forming, perhaps already populating a data room that an acquirer might wish to explore,” Hepworth said. “That's a very different mindset from spin-outs based on deep technologies or life sciences, where that journey could take 20 years.”

 

Elsewhere in the Ecosystem...

• Adsilico, a 2022 spin-out from Leeds University has raised £3.5 million to advance its digital clinical trial system. The company uses generative AI to create synthetic populations that can be used to carry out early clinical trials for medical devices, replacing animal trials and augmenting human trials. Regional venture fund Northern Gritstone put in £2 million and Parkwalk Advisors £1.5 million.
• Three start-ups were honoured in the 2024 European Institute of Innovation and Technology Awards. The changemaker award went to Mohamed Elamir, the co-founder of Aalto University spin-out Woamy, which creates cellulose-based biofoams as an alternative to plastic foams. The innovation team award went to Uppsala University spin-out Altris, which specialises in sodium-ion batteries. And the venture award went to Portuguese start-up Enline, which has developed a sensorless digital twin for use in power transmission asset monitoring, for example to anticipate vegetation obstructions and landslides.

Dutch start-up Scinvivo closed a €4.7 million series A round that will support the first clinical trials of its optical coherence tomography cancer diagnosis platform, along with work towards regulatory approval.