Dennemeyer Octimine

Technology Exploration / Licensee search

Constantin BahlmannConstantin Bahlmann

Written by Dr. Marc Feron, former EPO patent examiner.

Do not reinvent the wheel!

For any technology, quickly find out with octimine which companies are innovation leaders in which markets, how they do it, how they could help you, how old their technology is, the structure of their patent portfolio, whether including their technologies in your products may lead to IP disputes in certain countries, etc...

Find licensee


Whether to develop in-house or to seek an association with a supplier is fraught with uncertainty and danger, especially in cases the product or process being developed is to be marketed worldwide.

  1. If the contemplated development belongs both to the core business and to the development strategy of your firm, then a series of octimine FtO analyses may be helpful: please refer to the relevant octimine FtO application notes. However, since prudent manufacturers maintain a patent portfolio commensurate with their market share, read on to know how to find out which competitors seem about to enter your markets and how to pin hard numbers on their ambitions
  2. When seeking association with a potential supplier of relatively unfamiliar technologies, many parameters are important (price, reliability, quality, etc.) but also three kinds of information are essential:
    • Is this the best technology?
    • Is this technology part of the core business of the supplier? Is it likely to remain so?
    • Would adopting this technology from this supplier and integrating it in our firm’s own products or processes be likely to cause license or market access problems in certain countries or potential markets of interest?
    For answering these questions, a closer look at the patent portfolios of potential suppliers and that of their competitors in the technologies of interest may be advisable, for instance by setting up the following table and analyzing its documents.

    Supplier 1 CPC/IPC Country A Country B
    Technology 1   Patent 1  
    Technology 2     Patent 2
    Technology 3      

    Supplier 2 > Country A Country B
    Technology 1 Patent 3  
    Technology 2 Patent 4 Patent 5
    Technology 3   Patent 6 & 7

    Supplier 3 > Country A Country B
    Technology 1 Patent 8 Patent 9
    Technology 2 Patent 10
    Technology 3  

    An Octisearch can help address these questions by quickly finding the relevant patent documents to populate this table.

Step 1: Problem definition

Identify the technology field to be explored. In our example, octimine is used by an electric bicycle manufacturer wishing to adapt or borrow inverter-based regenerative braking techniques (converting the kinetic energy of the vehicle back into electricity, so as not to waste braking energy)

Step 2: Obtaining material for starting/seeding the octisearch

Step 3: Start your octisearch, restrict if necessary, find “jewels”

  1. Octimine was actually designed to better handle longer texts, so octisearching just three words, even very specific, is not the best way to use it. Therefore, the first batch of octisearch results may be disappointing, but our goal at this early stage is to find somewhat relevant patent documents (Jewels!) to meaningfully iterate the Octisearch query and thereby enrich the search vocabulary.
    • WO20141 55720A1, see also corresponding English language family member US2016/0052399A1, looks good because it details recent regenerative braking electronics, but is not about bicycles of golf cars: because of the mention of air brakes, it may be more suited for electric trains. Let’s make a note of it, though. It may become useful.
  2. Disappointing octisearch results can almost always be drastically improved by further restriction, as taught in the octimine application notes for FtO and Prior Art searches. For instance, results concerning golf cars or bicycles are more easily found by applying the “OR” cutoff filter to the 1000 documents of the Octisearch to find only those including either the word “bicycle” or the word “golf”.
  3. After applying this bicycle/golf octisearch restriction, expanding the results list by clicking on the green "+" to see the abstracts and figures, we identify for instance the following “jewels” just by looking at the figures:
    • WO2011002482A1 Electric bicycle drive system with regenerative charging
    • US2009112383 (A1) Golf car with regenerative braking

Step 4: Iterate your Octisearch!

Octisearch either or both of the above patent numbers, for instance only WO2011002482A1 if you’re looking specifically for regenerative braking bicycles. This provides a fair number of relevant results.However, something is not entirely right: specifically, the “similarity density curve” is very steep and shows a knee of the curve occurring for a similarity score of about 350, with only two documents with scores over 450 above that knee, and not many details about inverter design in these two documents. Ideally, we would prefer a “fatter” peak of the “similarity density curve” above the knee, including at least 10 to 20 documents well above it, and we’ll have to adapt the query accordingly if we wish to obtain results better focusing on the details we were expecting.

Step 5: Reiterate and re-reiterate your octisearch and results analysis!

(Octimine’s first law: always try to query octimine with documents, or preferably agglomeration of documents, which globally look like what you are looking for)

Octimine’s second law: Iteratively refocus your octisearch rolling snowball, repeatedly integrating your best-fitting patent number results in the next multiple patent number octisearch query!

Let us now try a multiple patent number octisearch using what we found: the initial patent number WO2011002482A1 (for its focus on bicycles), US2009112383 (for its focus on golf cars) together with the above-mentioned US2016052399A1, which has the advantage of focusing clearly on what we want, even if not exactly for bicycles, i.e. a modern inverter circuit set up for regenerative braking. Our goal remains to know more about the electronic circuits to be used in those low-cost, low power, lightweight applications on bicycles (no train inverters/variable frequency drives desired!). We now get a similarity curve with a less pronounced knee (at a higher similarity index of about 420) and a much fatter peak above that knee, which is a good sign:

Among the results, which are still quite car-oriented, we notice EP2711231A2 which has lots of specific electric bicycle braking details, and which therefore would be a great “seed” for an octisearch on electric bicycles.

Let’s keep our patent snowball rolling by adding this publication number to our multiple patent number octimine query (now comprising four relevant publication numbers separated by commas) and re-reiterate: the results now become more cycle-oriented, and the higher knee (around 480) and even fatter peak (meaning more very relevant results on our similarity curve now look even better!

We now notice in our list of results a number of jewels usable as octisearch seeds, i.e. numerous other documents concerning exclusively regenerative braking for bicycles (EP1415904A2, EP2481625A1,WO2012086459, WO2014200081 ) or electric motorcycles (EP2660093, EP2660096, EP2660097, WO2014064728 and WO2014064730 in Japanese; family members EP2910402, EP2910400) or electro-scooters (EP1488502). We also notice on the front page of the electric bicycle documents that those have many different IPC in the range B62 (with auxiliary electric motor B62M 6/40, B62K 11/00 engine-assisted cycles, B62L Brakes specially adapted for cycles), and this B62 IPC code helps quickly identifying documents relative to these cycle topics in the list of octimine results, without looking at the abstracts or figures.

We also notice that the range of vehicles in the results is still very diverse. But we just want bicycle documents! So as usual with octimine when we want better results, we re-re-re iterate, but this time discarding all non-bicycle documents from the query set and octisearching just the five most relevant bicycle regenerative braking documents which we found in the previous iterations.

Step 6: Re-Re-Reiterate your octisearch with your best documents!

Let’s now octisearch only the identified bicycle patent numbers of greatest interest, for instance EP1415904A2, EP2481625A1, WO2012086459, WO2014200081, reusing EP2711231A2.

Notice that in the above examples of multiple patent number searches, the addition of more patent documents in the query produced the expected effect: the peaks of the top IPC have remained on the regenerative braking IPCs where they were expected, which is good, with the advantage that the “similarity density curve” now is a lot less steep and shows a relatively unpronounced knee of the curve occurring for a similarity score of about 480, but this time with a fat peak including over sixty very relevant documents above that knee, quite a few of them concerning exclusively cycles (those with an IPC in the B62 range).

Step 7: Restrict your Octisearch for displaying usable statistics!

  1. When based on the standard 1000 document octisearches, the pie charts and other diagrams provided by octimine necessarily will be misleading. While the top 50 or 100 results can be expected to be quite relevant, the remaining 900 with much lower similarity scores would dominate and will pollute the results and pie charts, so they should be truncated/discarded
  2. In the present case, some documents among the top 100 are more related to cars than to cycles: a quick way of restricting the results to cycles is to use the filters to restrict the IPCs to the relevant IPC patent classification for Cycles (B62, see front page of relevant documents).
  3. This IPC restriction now yields 126 results in the IPC class B62, all with a relatively high similarity index and a similarity curve not having a pronounced knee, suggesting that all results are now statistically relevant. The set of results does not need to be truncated any further.
  4. The “similar documents list” now very relevant to regenerative braking for bicycles, reveals ca. 90% Japanese applicants; this is also reflected in the top applicants pie chart as follows: Honda and Shimano have ca 10% of patent families each; Sanyo, Yamaha and Kawasaki have ca. 5% of patents each.
  5. The results set can/should now be sequentially restricted to each of the main individual applicants by using the “applicant” filter, e.g. restricting to “Shimano” will show a list of 16 regenerative bicycle braking documents from the Shimano patent portfolio, for evaluating their respective specific regenerative braking technologies and the corresponding relevance of their patent portfolios for selecting a potential partner or monitoring a competitor. The corresponding technical fields of the technical improvements claimed in these Shimano documents can be precisely analyzed in the top IPC diagram without requiring an in depth technical analysis of the individual documents, allowing non-engineers to make sense and significance of the patent portfolio.

Step 8: Technology over time heat map; competitors strategies?

  1. The Technology over Time Heat Map indicates when and in which general technical fields, (these technical fields are obtained by predetermined combination and agglomeration of results from IPC fields) patent filings for the searched technology have been particularly intense.
  2. For the present case study (regenerative braking for bicycles) it is no surprise that the “transport” and “electrical machines” fields show substantial recent activity, not least due to the impressive development of tens of millions of electric bicycles throughout Asia (and to a much lesser extent in Europe), and corresponding patent activity.
  3. Prudent manufacturers typically try to maintain a number of active patents commensurate with their present or anticipated market share in that technical field for cross-licensing and cost reasons. For this reason an octisearch might be seen as a quick, cheap and very rough proxy for worldwide market share /market ambition surveys and the numbers of patent family members in different countries may be used to discover telltale signs of expansion ambitions in certain markets.
  4. For instance, the above patent authority graph reveals the absolute strategic importance of China, Taiwan and the US for Shimano; Note also that although the lower height of the bars for EP or DE authorities might suggest that the European market is less important, in fact all 16 patent documents have one or more European family member (either DE or EP or both), suggesting that the European market is of absolute strategic importance as well. This country/authority analysis, really reveals the enduring worldwide ambitions of Shimano in the field of regenerative electric bicycle braking since 1996, especially considering its time line of patent application filings in this field, revealing marked efforts every 7 years (2003,2010) to upgrade its patent protection.
  5. Thus octimine is not just usable by IP personnel, but should be also used by industry strategists and marketing specialists to monitor signs of future ambitions or expansions of the competition. Ask yourself why a company is applying for patents at great cost in a country where it has no market share: it is either (a) a patent troll or (b) is seeking a cross- licensing deal with established parties in that country, or (c) may be ready to enter that market, the number of patents giving an important clue as to the future expected market share it is planning. Either way, it may be worrying news for established manufacturers in that market.
  6. By clicking onto the individual technical fields and years of the Technology over Time Heat Map (in the present case it is particularly instructive to click on successive years of the “transport field”, especially for the years represented in darker colors, such as 2010 for Shimano, showing regenerative bicycle braking improvements mainly in IPC fields B60L, B62 L and M, and G06F), the other diagrams as well as the list of similar prior art are updated, displaying the main IPC codes and major applicants for that particular year, as well as the list of their patent filings, thereby providing a glimpse of the evolution of technology and of the main applicants in that technical field portion of the octisearched query. This octimine feature permits watching what competitors are up to in any specific technical field, and has the potential to quickly track their patent strategy over time in that particular field.
  7. Coming soon in octimine will be a heat map of IPC and CPC fields, not just imprecise general technical fields, which will visually represent in one glimpse where and when the action is or has been in those fields, but already now, by clicking through the heat map and looking at the evolution of the IPC codes over time, it is already possible to see this information: which IPC codes, i.e. which problems and solutions can be considered “hot” at any given time, and who are the major players in that field for each such solution.

Start your free patent search now.

+49 89 444 78663 0 | | © octimine technologies GmbH, 2017