UIDP26 puts the spotlight on collaboration

University-Industry Demonstration Partnership (UIDP), a membership organisation that facilitates partnerships between universities and industry, held its 26th conference (UIDP26) in April, welcoming hundreds of delegates to the Silicon Valley headquarters of networking equipment maker Cisco for the partnership’s largest gathering to date.

The first keynote speaker was Vivek Wadhwa, distinguished fellow and adjunct professor at Carnegie Mellon University, and co-author of The Driver in the Driverless Car: How Our Technology Choices Will Create the Future.

Wadhwa invited the audience to imagine two opposing scenarios – a utopian future such as the one depicted in Star Trek, and a dystopian future similar to the one shown in Mad Max, pointing out that, in fact, humanity was currently hurtling towards both of these and society would have to make a conscious choice which version it wanted.

The problem was exacerbated by the exponential speed of progress, Wadhwa said. It took thousands of years from the stone age through agriculture to modern-day technology, a reality that society seldom appreciated – illustrated, for example, by people getting upset that the latest iPhone did not boast all the features they dreamed of despite the release of a new iteration every year, each one more advanced than the last.

Technology needed to benefit humanity, argued Wadhwa, especially as smartphones were expected to reach the processing power of the human brain by 2023. But the intrinsic truth that each invention could be used for good or bad had always been true – even something as basic as fire could be used to warm a house or burn it down.

Trying to stop the development of robots and artificial intelligence would be futile, Wadhwa said. The technologies already existed and with a growing elderly population we would increasingly need to rely on robots in the care sector.

Other technologies were moving closer to daily reality – genomics, microbiome therapies, gene editing and solar energy – and all of them were more or less unstoppable, Wadhwa noted. They were unstoppable even in the face of political decisions, such as solar energy, which had experienced setbacks with the US Trump administration trying to force the use of coal but which was still expected to serve all humanity’s energy needs within 14 years, he claimed.

All of this would lead humanity to some difficult discussions within a decade as we would have to decide where to draw the line on each technology’s use. While Wadhwa did not offer an answer to this, the stark warning and enthusiasm in equal measure set the stage for three days of discussions around the benefits and dangers of technology.

Wadhwa’s point about political decisions also surfaced in other discussions. Pablo Debenedetti, dean for research at Princeton University, who came on stage with Kristen Wright, director of Cisco Research Centre and Open Innovation, explained that the Obama years were a spectacular run for the innovation ecosystem. When asked if he was optimistic about the Trump White House, his brief reply – “no” – led to an agreeing laugh from the audience.

University-industry collaboration needed to continue and Debenedetti pointed out that interacting with corporates led to scientifically interesting questions that would take much longer to formulate if institutions acted alone. Collaboration was not, therefore, merely a short-term way of getting funding and Debenedetti cautioned that approaching it as such would be a recipe for failure.

He explained that Princeton was in an enviable position when it came to collaborations – the university’s partners had included pharmaceutical firms Eli Lilly and Merck and chipmaker Intel. Many of the institution’s alumni had gone on to lead international companies – Meg Whitman, who led technology company HP and, after its split into two businesses, Hewlett Packard Enterprise, and Eric Schmidt, former CEO of internet company Google, as well as Jeff Bezos, chief executive of e-commerce and internet company Amazon.

Wright echoed Debenedetti’s argument that Cisco’s best innovations had come from collaborative research and the most successful cooperation had come from the corporate putting boots on the ground. Cisco would take a range of different approaches to partnerships, Wright continued, and would occasionally go as far as sponsoring a chair or leveraging its distributed network. Wright also noted that Cisco was aware of the lack of women in technology and was interested in identifying root causes – a topic that formed the basis of several other talks.

One female speaker who had succeeded in making a mark was Jennifer Rexford, professor of computer science at Princeton University, who spoke to delegates about her research.

Rexford explained that industry and interdisciplinary collaboration was important in computer sciences and noted she was currently working on networks that would be easier to manage – a key focus as the sector was facing advances such as autonomy and new rules around privacy. Rexford, who had joined Princeton from telecoms firm AT&T, rejoiced in the possibilities that academia had afforded her, noting that one of the most frustrating experiences in her previous job had been that researchers were forced to accept the existing network as a given rather than having the opportunity to build something more efficient.

Both the discussions around the need for new networks and people moving from industry into academia, or indeed the other way around, resurfaced as topics in other other panels, as did the discussion about gender equality.

The latter was taken on in a breakout session by Mindy Cohen, president of executive talent search agency Higher Talent, Dennis Fortner, director of corporate relations at Carnegie Mellon University, Gaylene Anderson, US director of contracts and alliance management at pharmaceutical firm Boehringer Ingelheim, Brad Fenwick, senior vice-president for global strategic alliances at Elsevier, a subsidiary of analytics company Relx, and Gert Lanckriet, principal applied scientist at Amazon Music.

One of the panel’s key conclusions was that one side did not have a full appreciation of the other side’s constraints and expectations when collaborating. They agreed there was a need for more people to make the switch to help each side understand the other better, though they also cautioned that there was a limit to how many people each side could take on before the disruption would become too big.

Fenwick meanwhile also took on the topic of women in science and technology together with Serpil Bayraktar, principal engineer at Cisco. Fenwick provided an overview of his company’s gender working group, which was set up to increase diversity on editorial boards and conference panels as well as to apply analytics to gender issues.

Fenwick made a range of striking points, most notably that the proportion of women among researchers – defined as the author of a paper for the purposes of Elsevier’s analysis – had risen substantially in the US, up from 9% in the 1996 to 2000 period to 40% today.

One finding that had come as a surprise to Fenwick and his team was that in the biomedical sector women actually published more than men. Additionally, the citation and download impact of women’s research was bigger than that of men.

While women were found to collaborate less internationally than their male colleagues, the rates for university-industry collaboration were roughly equal.

Bayraktar noted that pregnancy and maternity leave remained huge issues, as did sick days for children and picking them up from school. The last two aspects in particular were barely ever discussed, she said, when approaching the challenges.

The fact that women were forced by these external factors to be absent meant that promotions mor
e often than not came to male colleagues and compensation was not distributed equally.

Returning to technology, Ed Felten, professor of computer science and public affairs at Princeton and former deputy US chief technology officer for the Obama administration, joined Aaron Kleiner, director for industry assurance and policy advocacy at software developer Microsoft, on stage to talk about cybersecurity policy in the age of artificial intelligence and the internet of things. Chip Hay, who leads corporate engagement in the western and Pacific area for Princeton, moderated the talk.

Hay observed that the early days of the internet seemed like ancient history compared with what was possible nowadays.

Felten agreed but called out the “problem children” particularly in the internet of things sector – companies, he said, that employed worst practices such as hardcoding a password into their software, making it easy for malicious actors to break in because they could more easily guess or even see the password in the source code, or not authenticating updates, meaning hackers could introduce bad code by convincing the device it was a genuine update from the manufacturer.

Products such as smart thermostats were nice to have, Felten said, but consumers needed to understand that such devices also had the ability to extrapolate other information, for example which room was currently occupied.

The problem, according to Felten, was that devices were so cheap that companies felt they could not afford to build in security – “an argument that would not fly in other industries”, where it would be unthinkable, for example, to produce a lighter that easily caught fire. The US regulator was trying to fix the issue, investigating companies that breached best practices, but their tools and powers were limited.

Kleiner added that the EU’s General Data Protection Regulation, which came into force in May, was likely to prove a game-changer for any business that wanted to do business in Europe. Microsoft had also taken other steps to help the ecosystem, introducing a secure open-source and end-to-end internet-of-things platform dubbed Azure Sphere.

Felten explained that working towards standardisation and frameworks was important, and Princeton had set up a house filled with smart devices to understand better how someone might try to break in.

The issues around internet-of-things technologies were also picked up on by Vyas Sekar, associate professor at Carnegie Mellon University, and Jim Warren, program manager at Cisco, in a breakout session. Sekar gave a noteworthy example of how researchers managed to break into a house by hacking into a smart plug to turn on a toaster, which in turned activated the fire alarm, resulting in the windows opening.

One of the key challenges faced by the sector, Sekar said, was that there were hundreds of operating systems – each vendor tended to develop their own – making it much more difficult to plug security holes. Traditionally, developers had largely been faced only with Windows, Mac and Linux for desktop machines, and Android and iOS for smartphones, Sekar explained.

He also noted that smart devices could be used to cause more widespread damage. In October 2016, he reminded the audience, hackers had broken into thousands of smart devices, such as fridges, to bring down the internet on the US east coast for the majority of a day.

But it was not all doom and gloom, he reassured the audience. Sekar’s lab was working on generating intellectual property that would be released under an open-source licence to help companies create a secure and privacy-respecting internet of things. One of his team’s approaches was to create a context-aware network, so that if a toaster was set off but there was nobody in the house, the system would know that it was a malicious attack from outside.

Warren, meanwhile, explained how Cisco was working on fog computing – a type of computing that shifted processes across multiple layers and enabled systems to react locally rather than relying on a central application in the cloud. The technology could be used in security monitoring, where local nodes would analyse footage rather than overloading the network with large video uploads.

Taking up Sekar’s point about the toaster scenario, Warren noted that fog computing could also be used to authenticate devices locally so that access between devices that had no reason for being connected, such as a thermostat and a television, could be prevented.

Another area that received a lot of interest during UIDP26 was healthcare, the subject for several panellists and keynote speakers.

One of these speakers was Gaylene Anderson of Boehringer Ingelheim, who gave an overview of how her company approached collaborations.

She explained that Boehringer Ingelheim employed scouts by indication rather than by geography. The company had publicised on its website details about how to contact relevant scouts and Anderson invited delegates to get in touch through those channels.

Boehringer Ingelheim had historically been focused on Europe, Anderson said, but was expanding its operations in the US. The company had also launched a “research beyond borders” program, which focused on anything that was not a core area for Boehringer Ingelheim, such as regenerative medicine.

A panel featuring Kristen Wright, who had earlier talked about gender equality, as well as Marlon Harvey, global solutions portfolio leader, healthcare architect, at Cisco, Kathleen Richard, co-lead of co-invention program Optum Cisco Clinic of the Future Centre, and Alex Gao, senior director, head of Digital Health Lab at Samsung Research America, the research and consultancy subsidiary of electronics producer Samsung, took on technology in healthcare.

The panel looked at issues around privacy and security related to healthcare data, such as the Health Insurance Portability and Accountability Act (Hipaa), the US law governing how medical information should be handled.

Gao noted that companies whose business model was data mining would inevitably face issues. That also made it difficult, Gao argued, to mine healthcare data without compromising user privacy or security.

Wright took a slightly more optimistic stance, saying patients might in future elect to share their data, which in turn could lead to new business models. The consumerisation of healthcare, Wright observed, was clearly happening, underlining that “we are all patients”.

Richard added that Hipaa mandated strict security, but strict security could force a company to look at data to make sure it could not be tampered with. Looking at data for this purpose could already constitute a breach of the law and, Richard said, policy would need to address this tension sooner rather than later.

Finally, Rob Lowe, chief executive of innovation management software producer Wellspring Worldwide, looked at how patterns in partnerships illustrated trends in technology strategies.

Lowe said the global annual market for R&D was more than $2 trillion, dwarfing everything else – even alcohol and tobacco were only $120bn. Asia already represented 42% of R&D spend, with Japan being the number-one patenting player on the continent ahead of, in this order, China, South Korea, Israel and India.

Graphene research had proven particularly fruitful, with an exponential growth of research in that area and heavy investment since 2010, when Andre Geim and Konstantin Novoselov jointly received the Nobel prize for physics for their ground-breaking experiments with the material.

Interestingly, Lowe said, corporates had taken very different approaches to collaborative research. Contract manufacturer Foxconn was almost exclusively working with Tsinghua University on graphene technologies, while electronics company Samsung had spread its partnerships across multiple institutions.

Stanford Uni
versity, meanwhile, conducted 70% of its collaborative research into fuel cells with car manufacturer Honda.

One particularly interesting corporate was Motorola, Lowe said, as nobody he talked to ever predicted that it was this company, usually perceived simply as a smartphone manufacturer, that had the most partnerships in autonomous vehicles and was more connected in the ecosystem than anyone else.

The mix of cautious warnings and excitement for the future remained palpable throughout UIDP’s latest conference, both on stage and among delegates on the floor.

Opportunities for university-industry collaboration continue to abound – particularly as the internet of things is likely to disrupt most industries – and UIDP president Anthony Boccanfuso, who steered proceedings at the conference, was clearly ready to lead the charge. And with the number of attendees the highest to date, the sector is certain to become an increasingly important aspect of technology transfer.