By citing prior art or related technologies, patent citations form the 'river beds' that connect the old with unchartered territories, across patent publications and non-patent literature alike. These evolutionary paths that fork and reconnect, that grow into powerful streams and at other points subside, tell the story of technological advancement and the flow of knowledge.
Automated and comprehensive processes such as NPA™ patent analysis augment a patent search result by also considering directly or indirectly related patents, that our proprietary process deems relevant. These patent trees are then agglomerated to form large networks of inter-related patents. This process is extremely comprehensive and will pick up on patents, that even a sophisticated patent search would have overlooked. Our research consistently shows that between 20% and 30% of the most relevant patents would not have been picked up in the initial patent search based on words, phrases and IPC codes.
An inherent quality of NPA™ patent analysis is the visual and insightful mapping of technologies and patents that cluster together based on the many citations they may share with others. Quite often these are not patents of the same patent family, or even same patent owner, yet NPA™ patent analysis recognised their similarility. Upon zooming into the NPA™ maps, the use of color codes and labels visualling the age of the patents can be used to show the development of a technology.
Further insights can be obtained by looking at the direction of citations, or we describe as the 'knowledge flow'. A citation reference from a later patent to an earlier patent brings with it the possibility that the later inventor built on or benefited from the knowledge disclosed in the earlier patent. This has applications such as illustrating the development of the technology, finding prior art, and predicting potential patent infringements.
As an example of this, consider Figure 1 which shows the very most central patents from an analysis of 60,000 patents in the hybrid car field. In this diagram, node colors differentiate between patent owners, with a special focus on two key actors. Green patents are from the US hybrid drive-train developer, Paice Corporation, and red patents are from Toyota. The label on each patent refers to the overall dominance of a patent in the hybrid car field, and it’s resulting rank within the owners patent portfolio, as well as the year of publication. For example, the P1(2000) patent designation indicates that this was the highest ranked Paice patent with a publication date in 2000. In fact this patent was the highest ranked hybrid car patent in the whole study, followed by the second highest ranked Paice patent, P2.
Note that the P2(1994) patent filed by Paice Corporation has arrows pointing towards the T4 and T5 patents for Toyota, both published in 1998, suggesting an apparent technology flow. While it is probably impossible outside of a court of law to prove that technology did flow from Paice Corporation, Toyota’s hybrid cars have been found in the US to infringe the Paice patent P2,US patent 5,343,970. This suggests that NPA™ patent analysis can be used to predict possible patent infringements. To confirm these apparent technology flows a review process using traditional and subjective infringement analysis methods is advised.
Figure 1. Apparent technology flows at the center of the hybrid car patent set
Doris Spielthenner
