15 August 2012
He is a man with a mission. Mathieu Goudsmits, director Research at security software supplier Irdeto, can become genuinely enraged about the illegal spread of copyright protected work. Within Sentinels, his company Irdeto was a partner in the CREST project, where watermarks as a security mechanism were examined in closer detail. The company was also involved in Kindred Spirits, which mainly considered the privacy of the user.
Irdeto works on the security of software in general. That could be the security of personal data in apps for tablets or in smart cards for pay-as-you-view television or for watermarks in videos to counteract illegal copying. In both areas the research department of the company participated in Sentinels’ projects under the leadership of Goudsmits.
Ing. Mathieu Goudsmits, Irdeto
‘Within CREST we investigated watermarks, especially for traitor-tracing applications. Pay-per-view videos, for example, contain watermarks so that you cannot copy them.’ Smart pirates, however, still manage to circumnavigate these occasionally. ‘You see that pirates sometimes place two videos on top of each other to camouflage the watermark. Within the Sentinels project we wanted to develop methods that will allow us to trace which videos these were so that we can track down the perpetrators.’
Link with universities important
Goudsmits is really positive about CREST. ‘The project was very successful. Several very good students did their final internship with us and a researcher on the project is once again working for us in-house for a period of several months.’ Such a physical collaboration with academic researchers works well, says Goudsmits. ‘At the universities, people often think too academically. If students experience the developments in industry for several months then that has added value for both them and us. Students then gain a more down to earth perspective. And we are too small to do pure fundamental research and so we value the link with the universities.’
Within the second project he worked on, Kindred Spirits, the recommendations were examined. ‘That project also mainly concerned the work of professional manufacturers. In the future I expect that our focus will shift more towards the security of user-generated content. Groups will gain in interest in using a tool to provide security for themselves and their content.’
Goudsmits has fundamental objections to free access to content. ‘Calls for more openness to protect people from large owners of films et cetera... people will move away from that idea in the future. The more user-generated content appears, the more people themselves will want security. If you throw everything open then eventually you only damage yourself.’ The security of small mobile devices such as tablets, e-books, and smartphones, are now the most important challenge in his opinion. ‘A growing amount of content is being stolen and shared. There are people walking around with USB sticks containing 12,000 books... quite ridiculous.’
Source of people
Sentinels is a good programme in his opinion. ‘The projects are making good progress and the cooperation with universities is really good. Such intensive collaboration leads to mutual understanding.
We have two important reasons for participating in these projects. The first, of course, is the knowledge generated: we only participate in those projects where we can see the intrinsic benefits for ourselves. However, a second main objective for us is to build up our reputation among potential new employees. We need good engineering graduates and PhDs. With Sentinels we are building a relationship so that we have a pool of people that we can draw on when we need to.’
Nevertheless Goudsmits has several points for improvement. ‘One difficulty with subsidies in general is how you arrange the intellectual property rights. Within Sentinels that was even more complex still because the decision had been taken to award funding not just to the academic institutions but, to a lesser extent, to industry as well. Part of the work and the funding was given to industry yet strangely the university automatically acquired all of the rights to the knowledge. I think there were some shortcomings there.’ Furthermore, academic institutions and funding bodies often have too rosy a picture about the utility and value of the knowledge generated, he says. ‘They have no idea what it still costs to convert a good idea into a product. As a company we often have to invest several times more than the original subsidy before we can actually make any money.’
Furthermore, Goudsmits is in favour of a limited number of partners per project. ‘Many subsidy programmes have the requirement that one of each type of entity must be involved: a large company, a small company, something international... that makes it impossible to you to focus the research. Furthermore, in such cases there are often partners who only participate to get the subsidy but are not involved at all in terms of content. This leads to projects which deliver knowledge that is not really relevant to anybody and which subsequently disappears onto a shelf. It is better to have a project like Sentinels in which just a few companies can participate and where these companies genuinely want to invest both time and money. Then you have the largest commitment, a longer and clearer focus and the best chance of a product.’
Photo: Sjoerd van der Hucht Fotografie
Text: Sonja Knols, IngenieuSe