OSS Watch team blog

A workshop run on December 7 2009.

Introductions

More and more researchers in the UK understand the concept of open innovation and more and more universities say that they support it. A large percentage of them even have a policy to adopt open source software.Even so, the number of open innovation projects running in UK institutions is still low. More clearly needs to be done - both to persuade people to adopt open innovation policies and to enable them to do so.

This workshop was designed to show how open innovation works in practice and how open innovation concepts can be applied to software development. In providing successful working examples of open innovation it also had the side benefit of making a good case for its application. (Although during the course of the day quite a few interesting problems and issues relating to open innovation were also teased out.)

Ross GardlerRoss Gardler from OSS watch introduced the day with a brief overview of the work carried out by OSS Watch in running workshops, providing speakers and expertise and advice: “We don’t make decisions - we help others make informed decisions.”

What is open innovation?

Ross also helped define the terms of the day by quickly explaining the principles of open innovation and citing the JISC description of a system that: “involves opening up parts of the development lifecycles for research, teaching and knowledge exchange .. external to the institution… in order to bring new perspectives, resources and collaborative opportunities.”Once these terms were laid down (and after Ross had given the most famous example of open innovation at work - the internet) we moved on to the first talk by Tim Francis from Pera.

How and where does open innovation work?

Tim built on Ross’ concise definitions of open innovation by giving Henry Chesbrough’s famous statement from 2003 that: “Open innovation is the use of purposive inflows and outflows of knowledge to accelerate internal innovation, and expand the markets for external use of innovation, respectively…”He followed with examples of companies that have made use of open innovation practice. These included: InnovationXchange, Innocentive, and Spread Shirt. The common strand tying these examples together – aside from their adoption of open innovation – was their rapid growth and success.Yet, Tim said, there are also serious issues of trust, problems relating to overcoming cultural differences between companies and academia and problems relating to feedback mechanisms within organisations that have tried to adopt open innovation.He cited a report that Pera carried out for JISC earlier this year, which showed that currently “there’s a lot of dissatisfaction.”

Open innovation in the Apache Software Foundation

Fortunately, the next speaker, Bertrand Delacretaz a senior developer at Day Software www.day.com and a member of the Apache Software Foundation was able to declare himself pleased with the way open innovation has helped his work.

The Apache web server, he says, powers 60% of servers around the world “and it relies on open processes and open code and quite a bit of crowd-sourcing.”Bertrand says that a good bit of Apache’s success can be attributed to the quality of the feedback that everyone working on it gets when they have something to add to the programme.

He then gave a detailed description of the feedback processes that the Apache developers use, including mailing lists alerting people when code is committed to the programme’s central repository, events issue trackers, email discussions and automated tests.“It’s really useful,” he said. “It helps develop a shared vision. There’s no scattered information. All decisions are made through mailing list. They get real-time status updates. They can do real-time help requests. It also helps create self-service archives. And the beauty of open-source is that you can do this in collaboration with your users.”

Emerging themes

During the course of this talk (Bertrand has helpfully posted the slides for this on his blog) one of the major themes of the day also began to solidify – that open innovation is not about chaos and throwing away control of projects.

Apache is carefully managed and there are numerous procedures in place to ensure projects have momentum and integrity. So, for instance, new developments will generally have an incubation period during which their direction is determined by much smaller groups of people before they are opened out to the wider community – and there are many safety nets when it comes to inserting new code into programmes.Also emerging by this stage was the question of whether or not open innovation really encouraged new ideas, or whether innovation is actually more often down to a few individuals. Delegates had asked it in different ways of Tim, Bertrand and Ross. Bertrand now answered by saying that open innovation “allows ideas to blossom”. It provides those who have had them with the feedback they need to take them further – and practical help in doing so.

Open innovation in TexGen

Professor Andrew Long from the University of Nottingham gave another example of a successful open innovation project – TexGen.You can get good details about how TexGen works here, but Professor Long also provided some interesting history of its genesis, including the fact that he had to be persuaded to make it open source by some PHD students “of the sandals and Linux” persuasion.

The experience of open source at TexGen appears to have been largely positive. Many more people have downloaded it than would have done were it not freely available – and much more money has come into the project than would have done otherwise. Professor Long cited funding that came in from Unilever and Voith Paper Fabrics for further research, saying he was convinced they wouldn’t have contacted him were it not for the fact that TexGen was open source and the businesses could see what it involved  - and indeed find it more easily in the first place.Professor Long  - although happy to admit he’d like a swimming pool – noted here that money wasn’t the object. It was “a happy spin off of the process” that allowed him to employ more people and do further research. But this financial success does help illustrate one of the points Ross was keen to emphasise throughout the day – that “open innovation is not the opposite of commercialisation.”

Problems in implementing open innovation

After the presentations, delegates were presented with a self-assessment exercise (which you can find here) designed to test how open and sustainable their projects were. As these results show, few scored particularly highly. “Knowledge in our sector is a problem,” concluded Ross. There were also problems relating to legal issues and marketing.

Delegates fill in their surveys

Discussion and conclusions

Negative as all that was, delegates did at least all say that the exercise had been helpful and the questions had helped guide their thinking.

More praise for the day emerged during the concluding discussion session and there seemed to be a broad consensus that – as one delegate suggested: “there was a line throughout that shows that open innovation would improve research.”But serious questions were also raised about how to bring about the cultural change that will allow for more open innovation programmes, and how to allow for the recognition and career advancement that academics naturally desire as a result of their work.

The lack of good open source software suppliers was also highlighted. Again the counter-argument emerged that much science depends on individual inspiration rather than group collaboration.Ross actually agreed with this latter point and acknowledged that open innovation may not always be the best fit. “But in other areas it does work,” he said. “The big thing about changing culture is to do it and show that it works.”The workshop had certainly achieved this latter aim, judging from the feedback in the session, and my own conversations with delegates who told me they had been given “plenty to think about and play around with” , given “really interesting pointers” about ways to shape the management structure of open innovation projects “not to mention a bit of inspiration”  and “a lot of interesting approaches to the issue”

But it’s also clear that much more work needs to be done to maximise the potential of open innovation across academia and in software development.