Insights into ICT hardware networks in Quebec: A focus on industry-university networks

October 19, 2015

Catherine Beaudry (Professor, Department of Mathematical and Industrial Engineering) 

Melik Bouhadra (M.Sc.A Student, Innovation Management)

Many industrial sectors face new challenges from emerging economies, but none more so than the ICT hardware sector in western economies. Indeed, very few semiconductor-manufacturing plants exist outside of South-East Asia. As the vast bulk of the Canadian ICT sector increasingly lies in the software sector, services, digital media and games, one may ask whether there still is a place for the hardware sector in Canada. If so, can policy mechanisms such as the creation of Center(s) of Excellence for Commercialization and Research (CECR) help maintain and expand hardware design and manufacture in Canada?

From 2006 to 2011, while overall revenues of the ICT sector have grown by 13% in the Montreal region (see Figure 1), at the same time, the manufacturing sector has shrunk from 16% in 2011 to only 5,8% of total industry revenues in 2011.

Figure 1

Figure 1: ICT sector revenues (in G$) in the Greater Montreal Region (Source : TechnoMontreal)

By investigating the position of Quebec ICT hardware firms on the local and international scenes, both in terms of design and innovation (examining the Global Innovation Networks – GIN) and in terms of manufacturing capacity (hence touching on Global Production Networks – GPN), we hope to understand the strengths and weaknesses of the industry in order to propose improvements in practices and policies. Amongst the factors examined, research capacity and technology transfer, measured by the relationships between universities and firms, will be studied.

In order to study the university-industry relationships, we mapped the various collaboration links from the Natural Sciences and Engineering Research Council (NSERC) funding programs, for hardware related projects. The links connect the academic principal investigator (PI) with the industrial leader of the funded projects to characterise the university-industry collaboration network using 5-year moving windows over 1993-2013 period.

In Figures 2 to 4, the size of the circle representing every network actor is a function of the number of partnerships in which it is involved. In other words, it depends of the number of links connected to it. And, the number of partnerships between two players determines the size of the link between them. The different colors distinguish the collaboration clusters within the networks according to the locally central organisation, generally the university.

Figure 2

Figure 2: 1993-1997 NSERC network of hardware-related projects

Figure 3

Figure 3: 2000-2004 NSERC network of hardware-related projects

Figure 4

Figure 4: 2009-2013 NSERC network of hardware-related projects

To illustrate the evolution of the network, three distinct NSERC network snapshots are examined. The 1993-1997 NSERC network (Figure 2) shows Nortel Networks Corporation in a central position. During these years, the telecom company Nortel was financially healthy and very active in terms of research partnerships. In addition, the network is mainly composed of French speaking universities.

The collaboration web then progressed into a denser network, as depicted in Figure 3. Nortel is still central but starts loosing its dominant industrial-actor position. Players such as Bell Canada begin to occupy a more central position (i.e. a stronger presence). This period also sees the emergence of McGill University as the academic organisation with the most traction. While the most active link joins McGill University and Nortel Networks Corporation, for the University of Montreal, Bell Canada is the preferred partner.

A few years later, the 2009–2013 NSERC network has considerably evolved into an even more complex web of intertwined organisations (Figure 4). In this latest network, McGill University and Bell Canada occupy the most central positions. Moreover, the strong link between both these organisations illustrates how closely they work together on joint research projects, much more so than say University of Montreal and Bell Canada. Nortel Corporation networks, in turquoise, is slowly fading away, which is not surprising considering that the company filed for bankruptcy in 2009. Over the years, as the technology moved out of the lab and into the market, the importance in the network of Nortel, as an innovator and developer of the use of optic fibres for telecommunication purposes, has decreased in favour of Bell Canada, the user and implementer of such a technology.

The density of the network showed on Figure 4 leads us to believe that the Canadian government might have achieved its goal of university-industry linkages. However, the average amount per NSERC grants for ICT hardware projects has steadily decreased since 2005. Furthermore, the total amount spent on ICT hardware projects in 2013 was nearly the same as in 2004. With limited resources, policymakers have to choose between investing massively into specific technologies to provide a national competitive advantage and spreading the capital into more research initiatives to foster a vibrant ecosystem. Our work shall provide elements of a response to this questioning on innovation in the ICT hardware sector. Social network analysis allows us to characterise the different relationships between specific actors of Quebec’s ICT hardware cluster, which will then be further investigated in detailed field studies. Next steps include the construction of a multilayer network, adding inter-firm collaboration information and co-publication data between various academic and non-academic organisations in a systemic analysis of ICT hardware clusters.