Introduction
Technology Transfer is a phrase much used when discussing research. In its broadest sense it can be used to describe and movement of a technology or technique from one context to another, and indeed it is often used when discussing the movement of technologies from developed countries to the third world. However, its use in relation with scientific research is usually to denote the movement of technologies developed in research laboratories into the market place, and hence to describe the process whereby technology is commercially exploited. The particular flow of technologies which we are most concerned with here is from universities to companies, but it must be remembered that for many companies the most important transfers of technology will be internal and that the transition from corporate lab to corporate product is a more important one.
American Universities, in official policy at least, tend to view technology transfer as a matter of providing benefit to the public. MIT's policy and procedures document on Technology begins its discussion of commercial development as follows
It has long been acknowledged that the primary functions of a university are education, research and public service. It is in the context of public service that MIT supports efforts directed towards bringing the fruits of MIT research to public use and benefit.
In many cases, mere publication of research results will be sufficient to transfer MIT research to the public. In other case, it is necessary to encourage industry, by the granting of license rights, to invest its resources to develop products and processes for use by the public.
[MIT, page 13]
There is a very real sense in which all the mechanisms of research oriented interaction discussed previously, including sponsored research agreements and consultancy agreements, are motivated primarily by the transfer of technology. The assimilation of a new technology is likely to involve much more than the simple movement of an artefact from one party to another, and although the licensing of a piece of clearly defined or tangible intellectual property often plays an important part in such transfers its importance is easy to overstate. This chapter thus focuses on technology transfer in its broadest sense, including an examination of the different moments of technology which might be transferred and the importance of different formal and informal mechanisms for doing so.
'Technology licensing is a good thing, and I think it needs to be considered, but I don't think it's the primary mechanism of technology transfer. I think the primary mechanism for technology transfer is direct contact between researchers in similar fields where they exchange information and know-how.... It's normally exchanged freely on the basis that it's of benefit to both parties. It doesn't cost anyone anything.... it's an informant network.'
[Rosenberg b] investigates industrial innovation in detail, and comes to conclusions regarding technology within companies which amplify the importance of 'tacit' knowledge and the development of organisational capabilities through accumulation of scientific expertise which were noted in section 4.1.1.2. It suggests that the reason that inventions are primarily made in companies rather than universities is that the technologies and needs of industries are not known within universities - such inventions come from the interaction of market needs with research expertise and so companies develop what are essentially their own specific technologies.
Such technologies are specific because, in many industries, patentability is not a key consideration and expertise is crucial. The results of his questionnaires reveal that patents are most important in industries where chemical composition is important, such as the chemical, pharmaceutical and materials industries. Patents are also important where products are devices which can be easily copied.
However, in other fields the importance of patentability is negligible. He suggests that for products such as missiles and aircraft it is likely to cost an imitator around 75% of the original development cost to successfully copy a design. This is because learning details of the technology is crucial to the duplication of its functionality, so almost the same expertise is required to understand and copy the design of the product as to produce it in the first place. This very specific expertise is what makes the technology proprietary.
It follows that the transfusion of such technology is a product of the spread of expertise, rather than a matter of licensing a patent or reading an article in a journal. Informal networks of information exchange and personnel movement are far more important, and the main role of patents may be to guard against potential infringement suits from others.
I suspect that patent licensing... is basically the tip of the iceberg of technology trading and sharing, most of which does not involve explicit licensing.
[Rosenberg b]
It seems likely that the IT industry represents another field where the role of patents is firmly secondary to that of expertise.
Rosenberg's research also investigates the perceived importance of research, and university research, in different disciplines across a wide selection of different industries. His findings show that computer science research is viewed as crucial to a large number of industries. Fields such as chemistry and, especially, physics, displayed a high perceived importance but a obtained low marks for the relevance of university research, whilst academic computer science research scored just as highly as the field in general.
Some quantitative data on the importance of formal technology transfer as a motivation for research collaboration is provided by [Cokar], which is based on an examination of 35 university collaborations in one company. Of these collaborations, 49% were intended to involve transfer of both technology and knowledge, 34% to transfer knowledge alone and 17% to transfer technology alone. In 43% of collaborations the professor involved was known to the corporate collaborator before the project, and in a further 31% of cases they knew of them by reputation. 57% of projects were judged to have successfully transferred information, 31% to have transferred materials, devices and software and just 3% (i.e. 1 so this is probably not a very reliable figure) to have transferred 'technology' (presumably patentable). In as much as these results may be generalised, they suggest that formal technology transfer is an uncommon result and is very rarely the only planned outcome of a collaboration.
A more substantive evaluation comes from a survey by the US General Accounting Office quoted in [Feller]. 168 firms were asked their motivations for taking part in the NSF Research Centers scheme. Of a total of 15 suggested motivations, the production of patentable products came last. Selected results are shown below:
Motivation: % of sample citing 'Match between research interests' 89% 'Interest in "state of the art" research. 88% 'Leverage of research funding' 48% /\/\/\/\/\//\/\/\/\/\//\/\/\/\//\/\/\/\//\/\/ /\/\/\// \/\/ 'The production of patentable products' 8%
Figure 5-1: Selected Results from US GAO Survey of Participants in NSF Centers scheme
Today, attention is focused on 'innovation' rather than invention. Innovation is a continual process, involving feedback, and is considered in a business context, involving the successful deployment of products and processes as well as their development. Successful innovation involves the use of new technology, which may have been developed in-house or acquired from elsewhere. Jeff Solash suggested that there are factors in corporate culture which have mitigated against technology transfer.
If a company are offered a piece of technology, they will often send it through internal channels to an expert in R&D, together with a vague covering memo asking him to evaluate it. The natural reaction this causes is often one of suspicion and resentment, worry for his own job security and an instinctive rejection of the technology - 'Why don't you increase my funding instead?'. A switch is required away from thinking in terms of a 'Research and Development' department to a 'Technology Acquisition' department - 'the president of Technology Acquisition should be in the business of acquiring technology whether it comes from internal or external sources'. This department also needs to be better integrated with sales and marketing efforts.
A company may acquire new technologies in a number of ways. One traditional way has been the use of suppliers and contractors - the producers of automobiles or aeroplanes often gain the benefit of new technologies via their adoption by the manufacturers of components. Another traditional way has been to develop it within the company. A smaller company which has developed a novel technology may be acquired or brought into an alliance, or a group of companies may come together in a consortium to pool the costs of research.
One thing which companies seek to do is to involve the risks inherent in product development. Like all projects, there is a risk that a development programme designed to lead in time to a new product may fail - at early stages this may be due to technical problems, other companies may establish themselves in a stronger position with equivalent or better technology, the market may alter or management conflicts may prevent the successful conclusion of the project. A corporation therefore wishes to hold a broad 'portfolio' of projects in different stages of development, to introduce new products as rapidly as possible and to have as much flexibility as possible. Many of these considerations lead naturally to a consideration of joint research and development ventures and to the licensing of technology. [Bower].
The shift from technology development to technology acquisition has undoubtedly been accelerated by the restructuring of major corporations, which as discussed previously has in general caused closer integration between research and development activity and business divisions, and by the huge complexity and expense of new products in many industries.
'companies no longer think .... if they don't develop it internally then it's bad. Companies are recognising that with everything from legal work to menial work to R&D that they can piggy-back off R&D that is being funded by the government or they can cheaply, by funding MIT research, take advantage of all the MIT overhead and equipment and experience to take a technology and move it forwards.... companies will start to recognise that it's the same thing, by giving up a little bit of a big pie you have more than if you try to keep all of a small pie. That's the way to make money. Industry is looking to be flexible in its R&D and to capitalise on opportunities to collaborate, not only with universities but with other companies.'
'More than anything else, whether you really believe in it or whether you're cynical about it - it's more popular, it's the buzz. Licensing out is acceptable, it doesn't mean that you're weak - it means that you're nimble and competitive.... It's seen now as smart business. '
[see 9.1.6.1 - Interview with Alex Laats of the MIT Technology Licensing Office]
'People looking very much at what their core business is, considering whether to spend money internally or outsource, ultimately this must benefit the universities. The good news is that it's reached the top of the industrial agenda - when I started in this office seven years ago, one felt that, both in companies and universities, industrial liaison officers were somewhat marginal. Now, both are much more a part of the core business, and the quality and the exchange has increased dramatically, as has the support in both sectors.'
Small firms are unlikely to be able to take part in longer term research projects. As discussed previously, they are less likely to have either the resources or the long term focus needed to invest in such schemes, which from the point of view of the company involved in highly speculative. Other mechanisms of technology transfer are therefore more important for such companies. These include the kind of informal links discussed in 4.3.14.3.1. The employment of graduate students can be an important mechanism for such companies.
'Small local technology companies are increasingly employing graduates, and that's a very real, and in some ways the most important, source of technology transfer: the movement of intelligent people.'
Although, as discussed in 4.1.1.2, the progression from basic research to product is often far from linear, by the time that a piece of technology has been assessed as being of likely commercial value, and is under consideration for patenting, at least one commercial application must have been observed for it. Unfortunately for technology transfer offices, the state in which a university research programme is likely to have left the embryonic technology is not usually sufficiently advanced for a company to be able to easily adopt it and turn it into a viable product with a minimum of effort or risk. Alex Laats at MIT stressed the centrality of these problems to technology transfer.
'The real difficulty in terms of taking technologies and getting them from an embryonic stage to a product stage, and then from a product stage through a scale up to a large market is funding along each step of the process. Whether it be an industrial sponsor or the US government that is funding research here at MIT, MIT is not in the business of delivering products - its role tends to be more open-ended - taking the technology to a very embryonic stage'.
The first stage is research - 'MIT all the way', followed by development - generally also funded by the sponsor and carried out at MIT. However, the government rarely goes beyond these to the next stage - 'the creation of prototypes on a bench in the lab'. Still more rarely will it fund trial manufacturing of a product. Industry by contrast, is usually much more interested in an idea if it can be presented in the form of a manufacturable prototype which they can then investigate the scaling up and marketing of. 'The problem is in the gap between R&D and the things that industry is interested in capitalising on'.
[see 9.1.6.1 - Interview with Alex Laats of the MIT Technology Licensing Office]
Most technology transfer initiatives, such as the US Advanced Technology Program (see 2.1.2.3) are designed to help overcome this, by supporting the development of technologies until they reach a more commercially viable stage.
One of the attractions of an ongoing active research collaboration is that technology will often be transferred to the company as it is developed (at least in the informal sense - they may have to license IPR separately) and so transfer will be from an internal research group to an internal, product oriented product development group. Even in this case, however, the biggest hurdle which the project has to jump is not from university to industry but from research teams (university and industry) to product development teams. As Alvey Programme showed, even when research collaboration has been highly successful and ownership is clear, its commercialisation is far from automatic (see 2.2.2.3).