"Study on Science - Technology - Industry Network" - Analysis of Japanese companies' return to core business and absorption of new technologies - (Japanese)

This paper analyzes the changes in the core technology fields in corporate research and development (R&D) activities, the relationships between and among the different fields of technology, and so forth, focusing primarily on exploitation of patent data, which have finally been developed as a usable research resource in Japan. Although there are certain differences depending on the technology field, significant correlations are generally seen between the composition of patent applications by type of technology and the composition of patent registration by type of technology. In other words, it has been found that "selection and concentration" of technology fields have hardly occurred in the process from patent application to patent registration. This is, we find, a phenomenon that can be observed regardless of sector and time. Moreover, when we focus on core technology areas on a corporate level, rather than on a sectoral level, composition by type of technology remains same in patent applications and registrations, with no notable changes observed in the core technology areas. This indicates that the relationship observed between patent applications and registration on a sectoral level is applicable to analysis on a corporate level. Changes in each technology area are not an independent phenomenon. Instead, they are part of the continuing process of co-evolution in various technology areas. That is, each technology area changes through interaction with other technology areas; these technologies influence one another both within and beyond organizational boundaries. We analyze the relationship among different technology areas, using several new methodologies and focusing on the long-term relationships among different technology areas within a company. Specifically, we demonstrate that the cluster analysis of technology domains within a company and the analysis of relationships among these domains can be done using the concept of IPC Co-occurrence (i.e., by examining the extent to which a single patent is assigned to more than one International Patent Classification code, indicating different technology areas). Then, focusing on a longer time span, we attempt to analyze technology trajectories within the company. With respect to Canon Inc., presented as a first case, we find that the company rapidly expanded its business, in principle, by building on existing core technologies, thereby incrementally developing new technologies in related technology areas, which leads to new business areas. We call such diversification of core technologies the "proximal diversification model." Meanwhile, regarding Takeda Pharmaceutical Co., a second case, we analyze changes in technology trajectories and the process of fusion of these technology trajectories. We find that the company, at least for some time after a new disruptive technology called genetic engineering came into being, intentionally implemented an information shield between the team in charge of new technologies and those in charge of the existing technologies. Then, after a sufficient period of time, the company integrated the newly developed technologies into the existing core technologies. This approach is believed to be effective in overcoming technology gaps. We call this the "venture insert model."