[Long Read] How Intermediate Research Institutions Can Bolster Semiconductor Manufacturing In India

In the year 2021, I detailed in two parts (Part-1 & Part-2), the landscape and possible avenues for Semiconductor Ecosystem in India. A one-line pitch for the uninitiated reader: In 2012, the Semiconductor industry was estimated to churn out an economic activity of over $7 trillion and was directly responsible for annual global GDP of $2.7 trillion.

In 2021, the industry clocked an outstanding growth of 25.6 per cent totaling to $553 billion and in 2022, it is projected to reach $600 billion more (eusemiconductors, 2021).

Hence, it is just too big an industry to ignore for the citizenry of the largest democracy of the world.

The recent policy dispatch by Government of India on Semiconductor Manufacturing includes a bullet on the future of SCL. In this context, this article explores the seminal question of how intermediate research institutions around the world are conceived, modeled, marshalled to extract value, fuel innovation and industry.

An understanding in how these institutions work in different national contexts will help India, in strategizing policy implications for future avatars of SCL or any other such institutions that may be gestating under Indian Semiconductor Mission (ISM).

Electronics products have a long gestation period between invention and commercialization of an idea, which needs significant effort to scale up, achieve consistent quality and yield, tune and test possible application for platform technologies.

The policy challenge is to cradle this “exploratory development” phase in the commercialization lifecycle, which is neither backed by venture capital nor by companies. This phase is the threshold where “idea is engineered into a real world application”; it is where research outcomes need to be formalized, tested and refined to reach the market as a viable business.

These engineering activities are often less glamorous than activities of pure research or product development but require long lead times and are resource-intensive.

For an industrial ecosystem to emerge and flourish “knowledge exchange vehicles” are an important component of Science & Technology Policy. These vehicles are the research laboratories by which new knowledge is diffused throughout the system and it is where different organizations participate in the innovation process starting from new ideas with market potential to their translation into new or upgraded industrial processes and/or new improved product/s.

The confluence of these outcomes is highly constricted in a typical university department juggling pressures to publish and teach, turnover of research staff, non-existent project management and contractual disciplines among other impediments.

The intermediate research organizations (IRO) operate as conduits between universities and industries but are autonomous and funded through permutations of public and/or private resources. They are vital intermediaries catalyzing break-throughs, licensing Intellectual Property, creating spin-offs, disseminating know-how, pooling R & D infrastructure, skilling work-force, networking and labour-exchange activities.

A few noteworthy models are illustrated as follows:

The Fraunhofer Model

The Fraunhofer Society was founded in Bavaria, Germany in 1949 with initial focus in geological research, soon expanding into many disciplines with support of public procurement policies.

It relied on military expenditure contracts until 1968, when it was formally incorporated in the Federal research budgets. Today Fraunhofer institutes engage in applied research in a national context where the total R & D budget is $125 billion with Industry footing most of it (65 per cent), followed by state (30 per cent), foreign institutions (4.6 per cent) and Not-for-profit organizations (0.4 per cent).

The Society is positioned between basic research and commercial technology development for industry. Two-thirds of the research activity is funded through industry contracts and public funded research projects and the rest one-third by federal and regional government base funding.

It operates 75 institutes and research units through-out Germany employing 29,000 people, majority of whom are qualified scientists and engineers with a research budget of $3.25 billion.

The Fraunhofer model consists of three pillars, each contributing equal amounts to the organization’s finances: a) Research directly contracted by industry, b) Publicly funded research projects and c) Pre-competitive research funded with based funding from Federal and Regional budget.

It leverages new knowledge by patenting contract research results which solve a technical problem on the basis of existing know-how that has been produced with substantial contribution from Fraunhofer staff and are deemed to have potential commercial value.

In 2021, 600 patents were filed and 26 spin-off companies were set up. The Fraunhofer management is structured as follows: society members form a Senate that appoints an Executive Board and formulates strategy.

A separate Policy Committee supervises financial matters. A Scientific and Technical Advisory Board assists the Executive Board in decision-making, while external Boards of Trustees advise the Institutes, Each Institute is led by a Director and a Steering Committee with Directors having joint appointments at local Universities.

The Fraunhofer model has transformed over several decades moving away from privileged incremental innovation, shed some degree of organizational rigidity, evolve business and IP strategies, 3-5-year frontline themes, encourage higher degree of cooperation within regional innovation clusters, formation of Fraunhofer Technology Academy providing part-time masters courses, teaching programs and specialized seminars, constitute Fraunhofer Venture Group to support commercialization and spin-offs.

Currently, the Society holds equity in 86 companies for a total value of $10.3 million.

The IMEC Model

IMEC (Interuniversity MicroElectronics Centre) was founded as a non-profit organization in Belgium by the Flemish government in 1984. Based in Leuven, starting with an initial investment of $70 million and about 70 members of staff, IMEC has grown to be one of the largest independent centers for R & D in micro and nano-electronics in the world.

Its mission is to operate 3 to 10 years ahead of industrial needs and to create sustainable development of the local industrial base through spin-off creation, collaboration and training. The local context in which IMEC is set typifies the challenges of innovation systems in small countries.

Belgium has few public research centers of excellence (for ex. the Catholic University (KU) of Leuven) and limited local links with some major companies (ex. Philips) but still is counted as a rich ecology of research organizations than larger European countries.

A problem of critical mass became clear soon after the beginning of IMEC initiatives and operating at an international level emerged as the only effective strategy. To achieve this goal IMEC engaged in a variety of activities aimed to increase its international profile and attract talent from abroad.

Moreover, the view that IMEC should work as a program-driven institute coherently organized around forward-looking, multidisciplinary, open-ended and highly networked projects has been instrumental in its success. IMEC is organized in three main units covering Business Development, R & D Operations (with specialized subgroups) and Corporate Support.

IMEC’s CEO supervises activities with the aid of a Corporate Strategy and Strategic relations office, a HR department, selected Executive Advisors and a training division.

The initial research outlook and competence of IMEC had two principle strategies a) “More Moore” and “More than Moore” where the former is concentrated on technology node scaling, advancing process technologies and miniaturization, while the latter is focused on inter-technology convergence and integration (like compound semiconductors, 3D integration, biochips etc..).

Now, IMEC has rebranded these strategies into System platform and Technology platform.

IMEC’s core research activities are strategically positioned in formulating “research program tracks” emerging out of basic science with commercial value. These programs are pitched to all the leading industries around the world to enter “working groups” and research output generated is shared within the working group on the agreements termed as — “IMEC’s Industrial Affiliation Programs” (IIAP). IIAP enables industrial partners to embed in IMEC as resident researchers. These programs enable the study of prototypes, develop further and test at IMEC.

The unique selling proposition (USP) of this arrangement is the sharing of risk, resources and new knowledge generated via information-exchange. To participate, the companies pay a fee that qualifies them to non-exclusive and non-transferable rights to exploit the know-how of the program (IIAP).

When the results generated during R&D lead to patents, each firm that has contributed can choose to co-own the IP, even a partner that has not contributed but has an interested in using the know-how as an end-user can also avail on a non-exclusive and non-transferrable basis.

There is also provision for specific research activities that are not shared with other firms of the IIAP when firms can negotiate with IMEC for conducting proprietary research.

Similar to Fraunhofer group, IMEC formed a for-profit arm (FIDIMEC) now renamed imec.istart to enable spin-off and tech start-ups. When an idea is approved for incubation, an activity roadmap is drafted to move technology from an idea to a product, then a new legal entity is set up together with a team to take forward the business.

A CEO and business development capacity sourced from outside IMEC contribute to the development of a viable business plan for selected spin-off and start-up ventures. Imec.istart supports and manages incubation by investing in start-up and reinvesting revenues from start-up in new spin-offs and IMEC’s stock option plan.

Since its launch in 2011 until 2022, as many as 220 tech start-ups and spin-offs have transformed into sustainable enterprises. IMEC’s with a resource pool of 4,500 professionals generates a revenue totalling $768 million in 2020.

The ITRI Model

The Industrial Technology Research Institute has been a trail-blazer of industrial policy in the Far East. It has been credited with a pioneering role in the history of economic development in Taiwan and is still a model to emulate for developing countries. ITRI was founded in 1973, forged with the merger of three research-oriented organizations previously operating under the Ministry of Economic Affairs: The Union Industrial Research Laboratories, the Mining Research & Service Organization and the Metal Industrial Research Institute.

In its early years, the Institute’s revenues were entirely government contracts and with time contracts from industry grew slowly in number and volume. The growth of ITRI was inextricably linked with the development of the Taiwanese semiconductors industry in the mid 1970s. At that time the problem for the policy makers was to foster the emergence of a whole new sector in the absence of significant infrastructure and competence.

Universities might have provided a starting point but they were not considered as a suitable environment for commercialization processes.

The decision to transfer technology from abroad and to invest heavily in training through ITRI was crucial. ITRI was instrumental in the formation of UMC and TSMC, ITRI stuck to its role as a knowledge conduit creating an ecology of small and medium size firms managing acquisition, integration, development and organization of IP.

ITRI executes projects mainly in two categories namely: Technology development and Industrial service. The model has evolved starting from a centralized budgetary allocation of the 80s, where a quarter of the budget goes to advanced projects, a quarter to exploratory projects and the rest of the half to Fundamental R & D.

The emphasis is on cross disciplinary domains with ITRI as the nodal agency of execution. The selection of projects is both top-down (Planning Division of the ministry and Director’s office) and bottom-up (R & D Labs) involving an Advisory committee of management, consultants and international experts.

Nearly 80-90 per cent of all the Taiwanese companies have had contracts with ITRI demonstrates the key role of the center. IP protection by filing patents gets more priority than scientific publications. The average duration of an advanced project lasts 3 years in contrast with 4 years for exploratory projects.

With the turn of the century, similar to IMEC and Fraunhofer, ITRI, starting from a non-profit organization, went on to develop a VC/incubation arm to attract capital and retain revenue returns. It then realized the need for international cooperation with global leaders.

ITRI has successfully transitioned from a catch-up to risk-driven innovation and technology development organization.

ITRI has a strong network of CEOs and engineers due to labour mobility between its labs and industry with over 25 per cent annual R & D staff turnover.

It highlights the institution’s role in training engineers and entrepreneurs and recruitments are equally split between university and industry. An overseas visiting scheme with 2 or 3-year binding on return is also in place which helps in much needed knowledge acquisition and exchange.

Contours For An Indian Model Of Intermediate Research Organization (IRO)

All the above models and the likes of ETRI in South Korea, CEA-LETI in France, VTT in Finland follow a familiar trajectory starting from the discovery phase, going through development, expansion and have reached the consolidation phase.

Even though these organizations differ in their span, size, management, national policy context and route of industrial development, their trajectories show the obvious culmination to operate in international technology markets and compulsion for globalized R&D.

IROs do face similar challenges of periodic institutional renewal, quest for sweet-spot between short and long-term targets, IP management, modulating to policy expectations and camaraderie with university and industry, and most importantly the need for the role of government grants and procurement policies in the early stages of these organizations. This is by no means an exhaustive account on all possible models, but it is a sufficient sample to derive key attributes that should guide India as India sets its semiconductor sail to ride ‘global manufacturing’ waves.

The key attributes are the following: the quantum of investment, acute sectoral focus, incentive driven labor mobility, effective IP exploitation, management by domain experts, out-come based nodal role for IRO between academia and industry.

Technology acquisition can bootstrap a developing country like India to leapfrog in setting-up an industry, but to sustain that industry, a long-term public and private commitment with vision backed capital combined with an IRO pool of skilled labour to absorb, adapt, diffuse and exponentially innovate in new technologies will be critical.

The failure to anticipate this middle-income trap with no roadmap for graduating from lower to higher value-added activities using IROs as catalysts would be an opportunity lost.

The ASEAN (ex. Malaysia) experiences should serve as useful reminders of pitfalls, which India would be prudent to avoid. Even though ISRO is an excellent indigenous model to draw inspiration from, Semiconductor Industry is a different beast and the words of Morris Chang (founder of TSMC) are a good note to end on: