Kenedict Innovation Analytics

Toyota Releases Fuel Cell Patents: What’s On Offer?

Toyota recently announced that it will make 5,680 of its patents related to fuel cells available for royalty-free use by other car manufacturers. Much like Tesla’s move last year (when Elon Musk announced Tesla would make all its patents available for use to anyone), Toyota aims to accelerate Research & Development in next-generation vehicle manufacturing. What is Toyota actually making available? How are these technologies linked to each other, and how have fuel cell technology clusters evolved within Toyota over the years? Network analytics provides an answer.

Written by André Vermeij, Kenedict Innovation Analytics

Interactive visualization of the evolution of Toyota’s Fuel Cells Technology Network

To get a view on what Toyota is actually making available, let’s take a closer look at all the company’s currently published granted patents with regard to fuel cells. The bibliographic data on these patents (titles, abstracts, technology classifications, inventors, publication dates, etc.) allows us to construct networks of related patents by showing each patent as a node, with links between patents based on e.g. shared inventors or technology codes. The publication dates then allow us to map the evolution of Toyota’s fuel cell patent network over time.

The Data

If we look closely at the initial speech in which the announcement was made (found here), Bob Carter (Toyota’s Senior VP of Automotive Operations) indicates that the deal only includes patents on which Toyota has sole ownership. This means that any fuel cell patents which Toyota has applied for together with other companies, universities or research institutions are excluded. In compiling the datasets, we therefore excluded these as well. Patents which list a majority-owned subsidiary such as Aisin Seiki, Denso, Aichi Steel or Daihatsu as a co-applicant were included. In the end, this leads to a total dataset of 4,975 granted patents published under technology classifications related to fuel cells and hydrogen storage¹. The remaining 700 or so patents are most likely not published online yet.

Network Evolution

Let’s take a look at how Toyota’s fuel cell patent network has evolved over the years. In his speech, Bob Carter notes that Toyota has been investing in Hydrogen Fuel Cell R&D for the past twenty years. Starting in 1995, the visualization below depicts the evolution of the network using 2-year intervals. The last few years have been combined, since the number of published granted patents from 2013 and 2014 is low due to the time lag between a grant date and its online publication. Click the visualization to enlarge it:

toyota_evolution_wide

Colours reflect clusters of patents based on inventor co-authorship on these patents – nodes are sized based on their betweenness centrality, a measure which reflects in how far a node serves as a hub in the network, connecting various clusters. As expected, we clearly see a significant increase in activity over the past two decades, jumping from 48 patents in ’95-’96 to 1,461 patents during ’09-’10. This perspective also allows us to find out how fuel cell technology clusters emerged, evolved and declined within Toyota. From 1995 to 2002, the purple cluster dominates the networks, with adjacent clusters slowly emerging over the years. From 2003 onwards, the network grows vastly, with various other technology clusters connecting themselves to the purple cluster, which eventually fades away and makes place for other groups of interconnected technologies.

Gaining insights from interactive visuals

Of course, static visuals don’t do justice to the wealth of information which can be obtained from these patent networks. The application above includes all the above networks and allows you to gain insights into the actual contents of the clusters. Clicking a node shows more information, including a link to the European Patent Office to read patent abstracts, claims and descriptions. The buttons on top allow switching between different time periods. You can also open the app on a separate page here.

Using the app, we are able to find out exactly how Toyota’s fuel cell technology network evolved over time. The importance of a few key inventors is clearly notable here: for instance, try clicking a few nodes close to each other in a cluster to find out who is responsible for the connections between these patents. The connections between the coloured clusters are often quite sparse, resulting from just a few collaborating authors across technology areas. The patents at the intersection of multiple clusters can be viewed as cross-over technologies, since they came to being using knowledge residing in multiple knowledge areas.

The importance of key inventors becomes even more clear when taking a look at the network of inventors which is also accessible from the interactive visualizations above². A quick stroll through the network shows us that Naoki Hasegawa and Tomo Morimoto play key roles in the large cluster at the top. Yuichi Yagami is a key Toyota employee in the green cluster, while Manabu Kato serves as an intermediator between the green and blue clusters. This thus shows us the difference between multidisciplinary knowledge and domain knowledge; those inventors with connections bridging multiple clusters have a more multidisciplinary focus, while inventors with a key position and connections in one cluster can be perceived as domain experts in their area.

Concluding

The above shows that a network analytical view on Toyota’s fuel cell patents can provide insights which can’t be obtained from traditional patent analyses. The interconnectivity between technology clusters over the years is clearly visible, with a number of key inventors serving as knowledge hubs within Toyota’s fuel cells R&D divisions. The company’s move is a bold one – let’s hope this will indeed spur R&D activity in fuel cell technology in the years to come.

¹: CPC (Cooperative Patent Classification) codes Y02E60/50 (Fuel Cells; including sub-groups) and Y02E60/321 (Hydrogen Storage).

²: The networks only shows the largest interconnected component of inventors. The full inventor network consists of 1738 inventors wiith 5768 connections between them.

10 Comments
  1. Hello Andre, it’s really a great work. May I also ask for TOYOTA patent list you collected? Thanks.

    • Hello Felix,

      Thanks for your message. I sent the patent list to the e-mail address you provided when posting your comment.

      Cheers,

      André

  2. Hello Andre,
    I’m also struggling to find TOYOTA’s open patent list.
    I want to ask for TOYOTA patent list you collected, if you don’t mind.
    Thanks.

  3. Hello Andre,
    You did an amazing work, you abstract lots of TOYOTA’s patents to get 4957 patents. May I ask for the 4957 TOYOTA patent list you collected? And I am really queried about the missing 700 patents.

    • Hi ArPin,

      Thanks for your comment. I’ll send a list of all patent numbers to your e-mail address. The remaining patents were probably not published yet at the time of publication of this article, or are classified under CPC codes which are not directly related to Fuel Cells. Hope this helps!

      Cheers,

      André

  4. Andre, may I ask what network mapping software you have used for this analysis, and do you have any recommendations on which ones work well for this type of analysis?

  5. Doesn’t this clustering actually reflect the management of R&D projects? for instance the Red “break out group” in 2007-2008. It seems, from the graph, that that might actually be a different lab or research center making an “independent contribution” to a larger research agenda.

    • That might indeed be the case here, Hugh. Patents are connected whenever they share at least one inventor, so the red cluster in the ’07-’08 network results from a group of collaborating inventors who actually have not collaborated with any inventors residing in the other clusters during that time period. Chances are thus high that this is indeed a group of separate inventors working at e.g. a separate lab or research center. Thanks for your comment!

Leave a Reply

Kenedict’s mission

To significantly improve Research & Development organisations’ decision-making by ensuring day-to-day usage of a social network perspective in the areas of Open Innovation, Competitive Intelligence, Human Resource Management and R&D Policy.

Contact Kenedict

Phone: +31 61 55 00 888

E-mail: info@kenedict.com