Discussion of all things patent mapping and analytics.

  • Home
    Home This is where you can find all the blog posts throughout the site.
  • Categories
    Categories Displays a list of categories from this blog.
  • Tags
    Tags Displays a list of tags that have been used in the blog.
  • Login
    Login Login form

Patent landscaping in seconds - case study on "Seeing Machines" driver monitoring

Posted by on in Amberblog
  • Font size: Larger Smaller
  • Hits: 1894
  • Subscribe to this entry
  • Print

How long does a patent landscape take to do? 

I asked this question once of a specialist employed by a French multinational, and she said up 'up to one month'. And I am sure that she did a great job.

But what happens if you do not have a month? Maybe a client is about to walk in the door, all you know is their patent  numbers, and you are trying to get a feel for the area they are in?  You might want to know how good their patents are, and who their closest competitors are.

What then?

Imagine, for example, you are were asked to look at the innovative Australian company Seeing Machines, which produces driver monitoring technology. How might you do this? 


The traditional approach

Yes you could try to predict the relevant keywords or class codes, and then run a traditional patent search based on these. This would time to fine tune the query, and to make sure that you not missed any relevant patents.

But in terms of time, this would only be the beginning of the process - the real time would be going through the hundreds or possibly thousands of patents that a traditional patent search would uncover. And this is fine if you have the time to do this - but maybe you are looking for an answer faster than this. Or maybe cost is an issue - time spent on a traditional patent search is valuable time that could be spent on other activities. For some searcher, this time can add up to hundreds of dollars.

These activities can be outsourced but this takes time to arrange, and time to wait for the results to come back. Maybe you are looking for results sooner than this, and this too leads to addtional costs.


The efficient approach

By combining our search tool with free public databases, you can easily and very quickly explore the patent landscape.

The (unrelated to Ambercite) patent search tool Patentlens allows you to easily run owner searches. The query would look something like this:



Which produces a list of results like this. 



There are some useful analytical capabilities you can explore in this tool. But I think a patent landscape is all about knowing about the patents around a portfolio - this can be very revealing. And this can be easy to explore in Family Cluster Searching.

Specifically, the results can be exported into a spreadsheet, where the 4th column is a list of patent numbers.



These numbers can be copied and pasted into Family Cluster Searching, without any editing required.



Cluster Searching will report the patents searched, but in some case as a 'duplicate', as shown below. This duplicate means that the patent concerned is in the same family as an earlier patent in this list. In this way, we avoid running duplicate searches within the same family.



All patents in this list are shown with AmberScore values - patents can be sorted by this metric to deterrmine the most important familes


With the highest ranked family being that of DE60140067, which also includes US7043056 and other family members.


What the most similar patents found?

So that is a ranking of the patents searched. But what about the most similar patents found?

The top 20 of these patent families are shown below, and we can show up to the top 2000 most similar. But even from this top 20 list we can start to see the leading applicants in this area, what specifically they are filing patents, and when the prority dates were.



If you want to present the results in graphical form, it is easy to export the results into Excel, and to manipulate them within Excel or other presentation packages to create the impressive outputs you are looking for.



So, in a few short seconds, you can

  • Find and list the relevant Seeing Machines patents
  • Rank them by family importance
  • Find the most similar patents, who is filing them, and when they were filed

All of a sudden, the challenge of producing a patent landscape, that does more than simply show of list of patents in the portfolio, becomes very achievable.


Unknown vs known citations

Unlike other citation based searches, Cluster Searching shows both 'known' (already recognised as a forward or backward citation to any of the family members of the query patents) and 'unknown' citations - which have not been recognised as citaiton but which still may be relevant.

An example of an unknown citaiton is  US5293427, ranked  33rd in our list, and filed by Nissan with a priority year of 1990 for a Eye position detecting system and method therefor.


Are are results reliable?

No patent search is perfect, but the results in this case are based on the citations found against  45 patents in 18 families - and the unknown citations linked to these citations. This meand that there would have between 18 and 45 independent patent examination in this group, probably by the same number of examiners - and many searches for the citations of these citations.

While no search is perfect, by the time all of these independent experts have had a search, they shoud have picked up almost all of the relevant prior art.  This in turn would have led to the forwards citations from both the Seeing Machines portfolio and its prior art.

The collective wisdom of all of these searchers would have meant that while not perfect - this search would be pretty good...and we believe more than comparable to other types of patent searching.


Would you like to learn more?

For further information on this analyis, including the other search engines tested, please contact us. 



Rate this blog entry:

Mike first developed an interest in patent data when working as a research scientist, and deepened this interest when working as an IP manager which led to his role at Griffith Hack. Mike has published in the areas of chemical engineering, patent management, the value of patents and the use of patent data in in a wide range of publications and forums, including the international journals Les Nouvelles, and Managing Intellectual Property.