4iP News

Assessing the impact of Computer Implemented Inventions

25 September 2015

Fraunhofer researchers Rainer Frietsch and Peter Neuhäusler talk to 4iP Council about the need for sound data on patenting Computer Implemented Invention (CII).

What is a Computer Implemented Invention (CII)?

A CII is a technical invention containing software as one element which makes it computer implemented. This includes:

  • devices in which some or all processes are controlled by a microprocessor, which is controlled by means of software.
  • processes (methods) which are computerised, i.e. where software-controlled computer or microprocessor monitors regulate and/or control the process.
  • a subset is formed by software "as such", which is only patentable at the European Patent Office (EPO) if it contains a technical effect that goes beyond the normal interaction of computers and software.

Are CIIs on the increase?

CIIs are clearly on the rise. The absolute numbers and shares of CII filings have grown steadily since the mid 1990s (with two exceptions during the economic/financial crises in 2001 and between 2007 and 2009). In 2011, more than 35 per cent of all filings at the EPO were CII filings. The majority of the CII patents are filed by applicants from the USA and Japan, followed by Germany, France and Korea. China has greatly caught-up in recent years, but still ranks just ahead of the Netherlands, Sweden and the UK.

What is the main policy issue around CII?

There is divergence between pro- and opponents of the patent system with regard to patents for CII. Critics argue that a computer program is not an invention in the strict sense but a creative work that is to be excluded from patentability. They argue that copyright is the right tool for the protection of computer programs. Proponents, on the other hand emphasize the planning reliability of patent systems, the clarity of its rules and the resulting incentives for innovation.

This divergence is evident in many countries and has led to different practices for CII patenting at national and multi-national patent offices. While some harmonization has already taken place, substantial differences still remain, for example, between India and China, but also between the US and the EU and even within the EU.

What does your research seek to do and why?

Current European policy-making and also German policy-making focuses on the digitalisation of industry and products that are part of daily life. Keywords like Industry 4.0 or Key Enabling Technologies (KETs) (which contain micro- and nano-electronics) are guiding innovation in policy making across many countries.

“The European discussion – so far – has lacked a sound empirical basis”.

Yet, most of the political discussions as well as harmonisation attempts (at the European and national level, e.g. in Germany) with regard to CII have been completely lacking empirical evidence on the economic impact of computer implemented inventions and the potential threats related to abolishing of their patentability. Our research aims to close this gap and provide this urgently needed empirical evidence or, at least, some of it.

Our focus has been on the effective trends, activities, and competitive advantages and disadvantages of CII to national economies from all around the globe at the European Patent Office. We have been analysing the effective role that CII plays in the competitiveness of European Industry. In addition, our interest has included companies’ assessment of the current situation in Europe.

Has this been done before?

Similar, yet not as exhaustive, attempts at assessing the economic impact of CII patents have mostly been made for the U.S. market. Analysis for the U.S., however, cannot easily be transferred to the European market as the rules for patenting CII differ between the U.S. and the European Patent Office (in the U.S., software "as such" is patentable).

For Europe, empirical evidence on the economic impact of CII is scarce. Besides, the policy discussion in Europe and the U.S. is completely different. The U.S. has longer experience with CII and there exist a number of studies dealing with software patents, business methods and also computer-implemented inventions covering the U.S. The European discussion so far has lacked a sound empirical basis.

What methodology did you use?

A prerequisite to our deep analysis of CII patents has been, first of all, a methodology for identifying CII patents within the total number of EPO patents. We applied a combination of keywords related to CII, based on previous research (Xie und Miyazaki, 2013). We searched these keywords in the title, the abstract and the claims of the respective patents and checked their appropriateness. In some technology fields, for example Chemistry, CII hardly occur. We excluded these areas from our analysis. Our estimates of CII patents thus can be seen as lower-bound estimates. In a second phase of our study we identified CII patenting companies and surveyed them in comparison to a group of companies who did not file CII. In this survey, we asked for motivations to file CII patents and tried to find out how companies deal with patents as a protection tool for intellectual property with regard to CII.

The study focuses on Germany in relation to other countries, why is this so?

The initial starting point of the study was a cross-party initiative launched in the German Bundestag, which essentially would have had the effect of a strict restriction or even abolishment of the patentability of CII in Germany with probable impacts on the European system, if it had been ratified. As already stated above, however, there is barely any empirical evidence on the economic impact of CII within Europe. We thus aimed to provide scientific evidence upon which future political decisions with regard to CII patenting can be grounded. Since political discussions on this topic have also been held at the European level, an expansion and comparison with other countries also seemed appropriate.

Which sectors generate most CIIs and why, according to the study?

About 78 per cent of all patent filings by companies originate from the manufacturing sector. This is to be anticipated, as patents are a less common protection instrument in the service sector.

Within manufacturing, CII filings are quite spread across a number of sectors, and especially so in Germany compared to other countries. The majority of the CII filings come from the sector "Manufacture of computer, electronic and optical products".

About four per cent of CII filings originate from other sectors, e.g. mechanical engineering or the automotive industry but also chemistry and pharmaceutical industry. This is one of our main findings: CII are relevant in almost all industry sectors and not only relevant for software or ICT firms. CII are also key enabling competences for the whole manufacturing sector.

Does company size impact CII?

When differentiating the filings by SMEs and large enterprises, it can be found that the largest share of CII filings come from large firms. It is interesting to note that SME shares at the USPTO are lower than at the EPO. However, in both jurisdictions SME shares of CII are lower than their shares in the totals. We explain this by the lower experience of SMEs both with patenting in general and with the tacit knowledge necessary to file CII (as the European regulation is not that obvious to newcomers). As a consequence of these factors, SMEs are also more cautious about their chances of enforcing patents and keeping others from infringing their technology. This is particularly true for CII.

What do your findings mean in economic and social terms?

We have been able to show that computer-implemented inventions nowadays are in widespread use and even define the competitive edge in a number of particular sectors. These computer implemented inventions are by no means limited to large enterprises as small and medium-sized companies report a large number of such patents. In addition, we estimate substantial employment effects as a consequence of CII. In the manufacturing sector in Germany, nearly one million jobs were directly or indirectly dependent on computer implemented inventions in 2010. And this is by no means restricted to the electrical engineering or ICT industry. In mechanical engineering as well as in the automotive sector more than 100,000 jobs in Germany were directly or indirectly dependent on computer-implemented inventions in 2010.

What does the research lead you to conclude?

Analytical discussion on the current case law shows that a departure from the patentability of CII would be both legally and economically more than questionable. Companies need patents to maintain their freedom of action or to open up markets. One of the main results of our survey is that companies are satisfied with the current legal situation and neither wish a sharpening nor a facilitation of the patentability of computer implemented inventions.

Furthermore, other existing instruments, in particular copyrights, would not be adequate or sufficient to protect this kind of invention. The use of copyright would thus significantly reduce the motivations and incentive for investment in innovative R&D and would, we expect, result in a high negative impact on the competitiveness of German and European firms.

What advice would you offer policy-makers on CII?

A departure from the current system for patenting CII would lead to widespread disavowals and certainly significant structural changes in the economic system of Germany, but also in many other countries. However, we would argue for a harmonization of the rules with regard to CII patenting, at least within the EU. National patent offices still have different regulations. It is difficult for companies to navigate through these different regulatory systems. This is probably less of a problem for larger firms than it is for SMEs and especially start-up firms who, faced with increased (transaction) costs, may be prevented from patenting their computer implemented inventions.

How has the research been received since its release?

Our empirical results have been highly acknowledged by the patenting community and by policy-makers, in particular, and our method of identifying and analysing computer implemented inventions has attracted strong interest from academics as well as from practitioners. We will continue to distribute our findings to a broader audience as the most striking but also surprising result to most people is that one third of all patent filings are essentially CIIs. Very few people are aware of this.

Download the full research report