A methodology to build an initial R&D portfolio for industry-university cooperation

How to build an initial portfolio of R&D projects to start industry-university cooperation? This paper reports an attempt to answer this question. Particularly, it presents a methodology devised from a 18-month action-research program held between a large multinational automotive manufacturer and the engineering school of a top-ranked Brazilian university. The paper proposes a 3-stage process oriented by the construction of QFD-like correlation matrices. A first matrix is built in order to correlate industry product performances and product components. Clusters of high correlations are identified and analyzed in order to define broad research strands of interest. These are, then, correlated with engineering competences to highlight the main types of know-how that are associated with each of the identified research demands. University experts in these competences are finally interviewed in order to fill a roadmap-like layer with possible cooperative R&D deliverables over time. Techniques used for data collection, analysis and visualization in each stage are presented. Possibilities of customizing the process are suggested when appropriate. Finally, based on the results, more general theoretical and methodological reflections are proposed in order to help managers tackling the long-standing problem of setting a shared R&D agenda between industry and university.