Materials Modelling

Home » 2014 » January

Monthly Archives: January 2014

Industry interactions of the electronic structure research community in Europe

I recently carried out a survey on behalf of the Psi-k network of the European ab initio research community and the CECAM-UK-JCMaxwell Node. The full report can be accessed here, and below is an overview.

The report explores the interactions of the academic Psi-k community with industry and is based on a semi-quantitative survey and interviews of network members. The evidence is analysed in the context of a prior report on the economic impact of molecular modelling [i] as well as of a recent study into Science-to-Business (S-2-B) collaborations [ii] in general.

Pertinent findings of the economic impact report were that the dominant electronic structure method, Density Functional Theory (DFT), is the most widely accepted ‘molecular modelling’ method and that it has become established in the electronics industry. Also of significance are the more than average growth in the number of patents which include DFT, and the growing interest in the potential of modelling in a wider circle of researchers in industry.

The S-2-B study [ii] emphasized the key role of the Principal Investigator (PI) in establishing and maintaining a satisfactory relationship, and the importance to industry of ‘soft’ objectives relative to outcomes with hard metrics.

All Psi-k board, working group and advisory group members, a total of about 120 people were invited to take part in the study, and 40 people responded, representing more than 400 scientists from 33 different institutions in 12 European countries. While it is acknowledged that this group will to some extent pre-select those with industry collaborations, the result that 90% of respondents work with industry is still significant. Main industry sectors of the collaborators are materials, electronics, automotive and aerospace and software. Density functional theory is almost always used in industry collaborations but classical and higher level theory also feature strongly.

It was noted that the Psi-k network represents some of the most widely used electronic structure codes in the world.  In fact, all electronic structure codes available in the leading commercial packages originate from Europe and are used at a few hundred industrial sites worldwide.

Psi-k groups that work with industry collaborate on average with 2-3 companies, typically on a long term basis. It is interesting that small groups are just as likely to collaborate with industry as larger ones, and also with roughly the same number of companies. There is however a correlation between the number of collaborating companies and the number of alumni in industry positions, which is consistent with the observation of the S-2-B study that the role of the PI and the depth of the relationship are the dominant factors.

Considering the different forms of interactions, informal interactions dominated, followed by collaborative projects, consultancies and training. Collaborative projects were reported by 75% of respondents with on average one such project per team per year. Nearly 60% of respondents had consultancy and contract research projects, with an average of one such engagement per research team every 1-2 years.  Training was least frequent but still more than 40% of respondents had training interactions in the last three years.

The main drivers for industry to collaborate are seen to be the expertise of the PI and access to new ideas and insights. As measures of success, new insights dominate followed by achieving breakthroughs in R&D. On the other hand, despite a clear ROI, cost saving is not generally the driver for collaborations. Impact was often achieved by unveiling mechanisms that could explain observations on a fundamental level and that had previously not been known or properly understood. The new insights thereby helped to overcome long standing misconceptions, leading to a completely new way of thinking and research direction. Similarly, electronic structure calculations helped to scrutinize certain concepts or aspects of engineering models. Less frequently so far seems to be the determination of input parameters for these models. However, the ability of simulations to screen a large number of systems, which would be prohibitively expensive if done experimentally, also plays an important role.

The above evidence and mechanisms of success indicate that the Psi-k network is largely in line with S-2-B collaborations in general, for example in terms of the relationships, importance of PI and the typical ‘soft’ measures of success.

On the other hand we can also see significant opportunities for further improvement. There is sincere interest as well as unmet need in industry.  On the one hand, the gap between industry requirements and what can be delivered by today’s theories and simulations is widely acknowledged. On the other hand, there is plenty of evidence that important and impactful topics can be addressed with current methods. However it takes a lot of time, effort and translation skills to identify and act upon these. Despite some activities by the network to further the exchange with industrial research, there is still too little common ground in terms of interactions, interests and language to develop the personal relationships that were found to be crucial for engagements between academics and industry.

However, we see evidence of successful mechanisms that can be built upon. These include utilising multiscale modelling approaches as not only a scientific endeavour but also as an opportunity to build a bridge in terms of communication and relationships. Also, relationships with industry at the level of Ph.D. training seems to be an effective mechanisms not only to train scientists with the relevant skills and understanding but also to build long term relationships between the academic centres and industry. Similarly, centres of excellence that are by their nature set up with industry involvement provide visibility and help to build relationships, although with the proviso [ii] that the single investigator can be the critical determinant.


[i] Goldbeck, G. The economic impact of molecular modelling. Goldbeck Consulting Limited, Available via https://gerhardgoldbeck.wordpress.com/2012/07/10/the-economic-impact-of-molecular-modelling-of-chemicals-and-materials/ (2012).

[ii] Boehm, D. N. & Hogan, T. Science-to-Business collaborations: A science-to-business marketing perspective on scientific knowledge commercialization. Industrial Marketing Management 42, 564–579 (2013).