Page 39 -

5 Conclusions

This work has expanded existing methods of patent analysis, introduced novel methods
of Web presence analysis, and applied both to the empirical investigation of energy
innovation. Preliminary results confirm and provide quantification of the notion that
different energy technology areas exhibit unique patterns of invention and innovation.
Generally, all energy sectors exhibit relatively isolated collaboration networks compared
to the rich collaboration networks observed in other sectors (e.g., software and
pharmaceuticals). The macro factors shaping these more collaborative industries help us
understand the industrial, regulatory, and technical features that limit energy innovation
and result in policy suggestions to drive richer collaboration in energy.

For energy as a whole, our results confirm the general notion that policymakers should
encourage multidisciplinary collaboration, university and national laboratory
commercialization efforts, and tailored support for regional innovation clusters. These
insights may be extremely helpful in the development and implementation of two new
initiatives, the Energy Frontier Research Centers and the Energy Innovation Hubs. These
initiatives are focused on discrete technology questions with specific innovation targets.
The analytical tools now at our disposal could help evaluate what are important social,
sectoral, market, and geographic considerations. For example, the Hubs are envisioned
as “multidisciplinary, multi-investigator, multi-institutional integrated research centers …
each comprise[d of] a large set of investigators spanning science, engineering, and policy
disciplines focused on a single critical national need identified by the Department [of
Energy] [89].” With enough data resolution, the analytical tools described in this work
can help evaluate the proposed institutional arrangements and identify structures that can
improve the likelihood of success.

Importantly, our results indicate that the dynamics of energy innovation are
heterogeneous across sectors, and the relative benefits of a particular strategy are unlikely
to translate smoothly from one technology area to another. Policies to foster innovation
and the deployment of those innovations need to be tailored to each specific technology
and its revealed pattern of innovation. This capability is important as the DOE embarks
on a series of experiments to drive innovation across multiple sectors.

With further work, we expect to increase the size and interconnectedness of the energy
patents database, achieve higher resolution on social, geographic, and institutional
variables, and refine the statistical methods of investigating linkages between
morphologies of invention and the production of breakthroughs. Together, these
objectives will constitute a novel analytic and technical foundation for informing
innovation policymaking.

32