A company is an effective way to turn research into practical solutions
Professor Aarne Ranta on research collaboration with companies:
– It is always fun when research leads to something that can be utilized straight away, says Aarne Ranta. When utilizing, you need to test the research together with those who will use it.
– The most common way to test your research for users is in a regular collaborative research project. In step two, it feels quite natural to start a business. That way you come a little further and it is easier for other companies to have research collaboration with another company. It is easier to sign contracts and it is a competitive way to conduct useful research. This kind of reasoning made us start our company Digital Grammars in 2014, after a few years with various collaborations with companies in previous EU projects.
To start a company is as I see it the best solution for research collaboration with private companies
– But it is important to keep business activities and the research under the auspices of the University separate. Research a such must be free and unbound. I make sure to distinguish between everything that applies to the company in terms of equipment and my time. It is important be in one place at the time also mentally. Then it becomes quite easy to draw the line and possible conflicts of interest are avoided.
– An example of an interesting collaboration is a language tool that we built for a language called Z, where the system translates Z-code into English. In the project we could use knowledge from previous research. As a result the language tool has now come into use in Altran in the UK, when they bought the system. It was in many ways very different to work like this. The system was much larger than what we could have built if it was an ordinary research project. It takes a lot of time to develop something that big, there are many details about routines and functions that must be solved and which are not really research but advanced technology development. The commercial system should be able to handle thousands of pages of code in a few seconds.
– In a pure research project in academia, you do not usually build something as large-scale and detailed, there are simply no funders for it. But in a company we get the opportunity and the whole thing will also be an evaluation of the technology, how it scales up to practical purposes. It also became a very close collaboration. The ordering company stated their needs and how they wanted everything to work, and they were the people who had a good understanding of the technology around the solution. We as researchers had extensive experience of language tools and had to transform this expertise into a creative solution.