University collaboration spurs GE’s European R&D expansion

The earth-moving equipment is moving in at GE’s European Research Centre in Garching, near Munich, in preparation for a massive expansion at the site. Already home to 200 researchers and a further 70 students and other staff, the centre is set to raise capacity to perhaps 450 scientists and researchers.

Now, though, the centre has been so successful that it has run out of room. A new building – an investment of “tens of millions of dollars”, says Carlos Härtel, the centre’s managing director, speaking in an exclusive interview with Science|Business – will allow it to have critical mass in all the areas it wants to concentrate on. Such a big investment was not approved lightly, but the decision to go ahead with it clearly reflects the centre’s success so far, and confidence in its future.

Just across the road lies the Garching campus of the Technical University of Munich (TUM), the top German institute in the Shanghai listing of world universities. That closeness was a key reason behind GE’s decision to open the centre back in 2003, says Härtel. And it has been a key factor in the centre’s success.

The centre was founded in the expectation that GE would quickly find interesting domains to collaborate with TUM, says Härtel. Now the university is deeply embedded in the centre’s work programme, with strongpoints in compressor technologies, combustion, magnetic resonance imaging and automated manufacture of composite materials. 

Working with a technical university doesn’t just bring with it access to talented researchers. It also allows resource sharing. Härtel instances large electronic systems and specific machine tools that a research lab would need to use, but not often.

The centre is by not means completely oriented around TUM. In magnetic resonance imaging – GE is looking in particular at specific disease-focused applications – it works with university and hospital partners in the UK, France and elsewhere in Europe, for example.

Being geographically close to TUM matters. “The intensity of the collaboration is absolutely a function of proximity between the two partners,” says Härtel. “You couldn’t easily replicate that model with a university 500 miles away.” 

Togetherness

The centre and the university form teams around each project, taking joint responsibility, working together in all senses of the word – sitting together and eating in the same canteen. At any one time in a project there will be a TU person in GE’s labs or a GE person at the university. “There are also those that stay at the university,” says Härtel, “but you can always walk over. You just have to cross the street.”

In the beginning most of the joint working was small scale, with the centre and the university trying out various ways of working, but “most importantly, learning about each other”. That led to the identification of areas where collaboration seemed most promising. Progress in those early days was not as rapid as many had hoped. “Some people thought it was all about topics and funding,” says Härtel. “But it isn’t. It’s all about people.” 

For anyone looking to replicate the model, he recommends starting slowly and learning as you go. A history of collaboration has to be built up over years. There will be many small projects that don’t lead anywhere beyond a joint paper. But over and above that, you need to have the right relationship between academia and industry – and in some countries that’s better than in others. 

Proper collaboration also entails having research partners who can “stand on their own feet”, says Härtel, and who have a background in industrial research or technology development – thought leaders who can understand your business. And by research partners, he means primarily people, not institutions. 

Härtel has one final ingredient in his recipe for successful collaboration: “Do big things.” Something has to be at stake for both parties, he says. If GE were to withdraw, that would leave a significant gap for the university, “not just in funding but also in research”. Equally, if the university doesn’t deliver then “we are in trouble”, he says, because the joint projects are generally linked to the critical path of a planned GE development. 

So GE is not interested in “nice to have” projects where it does not know what might come from a collaboration. “We don’t do that at all,” says Härtel. “I think the university prefers that too, because it gives them a chance to prove their value. That creates a history of delivery.” And that, he says, is the ingredient for building a bigger programme next year.

Looking forward to Horizon 2020

Like many companies, GE found the Framework programmes problematic – particularly the funding arrangements, which restricted Commission support to 50 per cent of eligible costs. That restriction will not be there in Horizon 2020.

Typically, says Härtel, the eligible costs are lower than the real ones. That often left companies having to find more than 60 per cent of the costs of a Framework project – and companies carrying the lion’s share of the expense will favour lower-risk technology development over groundbreaking but higher-risk research. And it contrasted unfavourable with the situation in, for example, the US, where in many health and energy research programmes companies need to find only 10 or 20 per cent of a the cost.

Härtel also welcomes reductions in the number of partners a consortium will need to get Horizon 2020 funding. Large consortia inhibit long-range research, he says. They work better with projects that are near-term, but the further out the research, the harder it is to work out the value chain. 

He gives the hypothetical example of a new semiconductor material, where it may not be clear who would produce it, how it would be manufactured, how customers would use it, and so on. “In those cases you need small groups,” he says. And in some cases, single companies should be able to apply for funding.