Writing the Background and Detailed Description

Having made the drawings and the core/summary of the invention, the most important parts of the disclosure are done, and below I will give some general thoughts about the background and detailed description part.


I suggest you try to keep the background rather short. Try to only discuss the relevant background. Furthermore, in the background, also mention prior art (i.e. known technique) solutions to the problem (if there are any), and also mention potential problems with the prior art solution (too complex, too expensive, too power consuming etc.) so that you can argue for the need for your invention.

Detailed description

The detailed description part of the invention disclosure can be long and should – of course – be detailed. Here in the text you describe all the drawings in detail. Try to come up with as many possible solutions (embodiments) for the problem as possible. By showing, let us say, at least two explicit examples/ways for performing one step of the invention, the patent attorney then can use that information to make the scope of protection wider. Referring to the example flow chart, pseudo-claims in earlier posts, one can show two or more different methods for determining the compensation value, and the same for determination of the characteristic value.

I hope I now have given you some tips and tricks for being able to write good invention disclosures next time you have determined an invention. Good luck!


Core of the Invention – Pseudo-Claims

Next step in the writing of the invention disclosure is to summarize the core of the invention. Now when the flow chart and/or block diagram is drawn, one probably can condense the core of the invention based on the drawing. I used do that in terms of “pseudo-claims”. Claims in a patent or patent application, define, in technical terms, the extent, i.e. the scope, of the protection conferred by a patent, or the protection sought in a patent application.

Since I have no formal education in the noble art of writing formal patent text – I have learned patent creation the street-smart way ? – I will by no means say that I can write formal claims, but something this is intended to be close to actual claims. I believe this will help the patent attorney drafting the patent to easier understanding of the steps I believe is novel and inventive.

Taking the fictive flow chart example, I showed in the last blog update, the corresponding pseudo-claims could be something like

1 A method implemented in a unit for determining a characteristic value of a received signal, the method comprising the steps of

  • determining a set of values for a first and second received signal
  • computing average of values for first and second received signal
  • determining a compensation value
  • based on computed average value and determined compensation value obtaining a characteristic value of the received signal

2 1 and further where the compensation value is based on received signal to noise ratio

3 1 and characteristic value is further used in further xxx action of the received signal

Number 1 describes the main steps (corresponding to the boxes in the flow chart I believe was the novel and inventive steps, i.e. 100,110,130,140). Number 1 can be seen as a proposal of actions that should be covered in the main claim.

Number 2,3 describes further details (the other boxes in the flow chart (no 2, 120 and no 3, 150), and can be seen as sub claim proposal)

Now, when finalized the drawings and core of invention you have – I will say – done the majority of the invention disclosure and the next step is then to write the background and detailed description.

Start with the figures!

Say that you got a brilliant idea that you believe is novel and inventive. Now you need to write an invention disclosure, which is a confidential document for use by a company’s patent department, or by an external patent attorney, to determine whether patent protection should be sought for the described invention. But how to write such an invention disclosure?

I used to say, start with a flow chart (or block diagram). I have, during my professional career reviewed many invention disclosures with a huge amount of text; long background, long summary  and a long description, but it was quite hard to understand what the inventor believed to be novel and inventive.

By starting with a flow chart or block diagram (or other drawings depending on what type of invention) and pin pointing the “combination of boxes” in the flow chart, or block diagram that might be novel and inventive, and then writing the text document and highlight these parts, you will gain two things;

  1. It will be much easier for you to write the entire invention disclosure
  2. It will be much easier for the patent engineers/boards or patent attorneys to review the disclosure and by that determine whether patent protection should be sought on the invention.

Below I show an example based on a fictive invention:

You have realized that to determine an estimate of a characteristic of a received signal comprising of two received signal streams, you need to perform a compensation of a mean value estimate over the two received signal streams, where the compensation value is a function of the received signal to noise ratio. The characteristic value of the received signal is then used in the further processing of the received signal.

Then, the flow chart in your invention disclosure might look like the below example

(100) receive the first and second signal stream

(110) compute an average value over first and second received signal stream

(120) estimate signal to noise ratio, SNR

(130) determine compensation value based on SNR

(140) determine characteristic value of received signal based on average value and compensation value

(150) use characteristic value in further processing of the received signal

Here, (120) “estimate SNR” is not new, but I believe the combination of (100), (110), (130), (140) is the really novel and inventive thing. Adding further a block diagram over the different blocks involved with then further give insight over your idea.

Hence, the quote; “A picture is worth a thousand words” is valid also in the area of inventing and patent creation.


Preparing an invention disclosure

Now back to the “How-part”; How to document an inventive idea.

Say that you got a brilliant idea that you believe is novel and inventive. Now you need to describe the idea which is done in an invention disclosure or an invention disclosure report. An invention disclosure, or invention disclosure report, is a confidential document written by a scientist or engineer for use by a company’s patent department, or by an external patent attorney, to determine whether patent protection should be sought (or not) for the described invention.

The invention disclosure typically follows the structure of a regular patent application and should include

  • Background, prior art

What is the background of the problem? It should also include the closest known (by the inventor) technique (prior art), i.e. how the problem was solved before (if there was any solution at all….)

  • Problem with prior art (complex solution, expensive, high power consumption etc. etc.)
  • Short summary, core essence of the invention

In some few words, what is the invention about

  • Advantages with the invention

Advantages using the proposed invention over prior art

  • Detailed description

As it sounds, technical details for enabling the invention

  • Drawings

In the next few updates, I will reveal how to write the invention disclosure in an efficient way. Hang on ….!

On the value of inventors from an R&D company’s perspective

In the last post, I showed that around 30% of the patents created in large companies within the area of digital communication, such as Huawei, Ericsson, Qualcomm, Nokia etc. originates from the top 100 inventors. Hence a very large part of the patent portfolio is created of less than 1% of the R&D employees, and therefore these top-inventors will create a large value for the companies.

Below I will do some back-of-the envelope calculation of the value of top-inventors. I will focus on Ericsson, mainly due to that they invest a lot in R&D and patents, and have a lot of patents (42000+ ) and license their developed technologies in open standard, such as the cellular communication standard LTE (“4G”) to other players in the market, and have been quite successful in their patent licensing business during the last 10-15 years and have a quite significant patent licensing revenue stream.
All the calculations below are based on open sources and clever guesses, and patent statistics is based on search using lens.org.
According to the annual report , Ericsson made 10 BSEK in patent revenues 2016, and during the last 5 years the revenue have been 7-15 BSEK, so hereinafter we use 10 BSEK/per year in patent revenues. Furthermore, Ericsson claims to have 42000 granted patents world wide (I guess they mean valid, still paid). Using the 30% created by the top 100 inventors, we have that 12600 patents, or 3 BSEK/year of the patent revenue originates from the top-100 inventors. That is 30 MSEK per year o average (Looking at the patent distribution between top 100, it is very skew, meaning top 10 inventors produce approx. 3-8 times than the inventors #90-100, but for easiness we take the average) the top 100 inventors “create” in patent revenues for Ericsson per year! Of course, creating patents and keeping patents alive come with a cost, so everything is not a profit.

But how much of these 30 MSEK is pure profit? Let us do an estimate. According to the annual report, the R&D spending is around 34 BSEK/year and with 24000 employees that gives an average cost per employee of 1.4 MSEK. The top-inventors are senior persons and hence the cost for such a person is, say twice the average number, say 3 MSEK/year. Patent maintenance and cosst associated to patent development are included in the R&D numbers, but to be conservative, we add the cost of creating and maintaining patents to the 3 MSEK number. How much is that? When looking at the Swedish patent office, PRV, a typical cost for a patent family (including 10 patents) over a life length of 11 year is around 700kSEK rounded to, say 1 MSEK. Hence the cost per patent per year is 1M/10/10=10 kSEK, and on average, a top-100 inventor at Ericsson has contributed with 126 patents, which means 1.26 = 1.3 MSEK per year.
So; A top 100-inventor at Ericsson generates patents giving a revenue stream of 30 MSEK/year, while the yearly cost of one of the top 100-inventor is upper bounded by 3+1.3 =4.3 MSEK.
That is quite good margin, I will say. Again, as a suggestion to all companies out there investing heavily in Research and Development; keep your top-inventors happy and loyal to the company, they are extremely valuable.

Inventor Productivity in Digital Communication

In the last two blog updates I have discussed the most prolific inventors in the world. But how does it look when looking into specific companies? For instance, how large fraction of the patents in a company are created by the top-inventors? Let us do some quick studies based on open information that can be found on the internet.

I will restrict the back-of-the-envelope calculations made in this blog post to the top 10 patent applicants for EP patents during 2016 in the field of digital communication   (I will check the statistic for other areas in a later post). The companies are (in the order of amount of filed EP applications 2016); HUAWEI TECHNOLOGIES CO. LTD, L M ERICSSON AB, QUALCOMM, INC., NOKIA CORPORATION, ZTE CORPORATION, INTEL CORPORATION, SONY CORPORATION, LG GROUP, SAMSUNG GROUP, NEC CORPORATION.

All these companies are known to invest heavily in Research and Development and have as a strategy to invest a lot in inventions and patents used for licensing their own technology and by that receive a return of investment of the R&D spending. All the companies above, as far as I know, have at least more than 10000 employees working with R&D.

Furthermore, I have used the search engine lens.org, and restrict the search to granted EP patents, which publication date was 2000-01-01 and later (up to March 2017). Then I have checked how many of the granted EP patents that the top 100 inventors were co-inventors on (again according to the patent search engine lens.org). This gives an indication of how large fraction of the company’s total patent portfolio that is created by the top-100 inventors. Below you can find the result.

Company             #Granted EP       #Top

……………….                 00-17                    100 inventors

Huawei                3729                      1244      33%

Ericsson              8661                     2656      31%

Qualcomm         4246                     2308      54%

Nokia                    4842                     1466      30%

ZTE                          904                      403        45%

Intel                       1735                       515        30%

Sony                     7038                     1656      24%

LG                          4330                     1433      33%

Samsung             8159                     1555      19%

NEC                      3674                     990        27%

Total                     47314                   14226    30%

As can be seen the fraction is surprisingly close to 30% in all the companies (with some few exceptions) and is probably a result of the Lotka’s law (I have to check, but that will be in another blog post ?).

Anyhow, one can conclude that patent creation is not evenly spread over the entire R&D, but instead made by a smaller amount of people. As an example, Huawei has 80,000 R&D employees , and 1/3 of the patents originates from 0.12% (!) of the R&D employees. For the second applicant on the list, Ericsson, who claims to have 24,000 R&D employees , 1/3 of the patents originates from  0.42% of the R&D employees.

Hence, from the above one can conclude that the top inventors, which may be less than 1% of the total R & D  employees generate approximate 1/3rd of the total patent portfolio (and patent value) for these high-tech companies!

Finally, as a suggestion to all companies out there investing heavily in Research and Development; keep your top-inventors happy and loyal to the company, they are extremely valuable. In the next blog post I will do some back-of-the envelope-calculations for estimate the value of prolific inventors, so hang on …


Top-10 Inventors, Granted EP patents

Last update I gave a list of the Top-10 inventors when it comes to granted US patents. This information was based on the wiki page prolific inventors, but I have not found any similar top lists for other jurisdictions. However, the patent search engine lens.org has a great feature, where one can extract top 100 inventors for a specific search string. Since the search is on inventor names only, so if there are more than one inventor with same name, the result shown is the total for all with the name “Sven Svensson”, so the result might not be entirely correct, but at least it gives a good indication.

Anyhow this is the top-10 list as on (20/5-2017) for granted EP patents, based on lens.org search engine

  1. Focke Heinz                    (363)
  2. Silverbrook Kia             (330)
  3. Van Den Berg Karel    (287)
  4. Biedermann Lutz         (252)
  5. Gueret Jean-louis         (231)
  6. Suzuki Takashi               (224)
  7. Mueller Bernd                (213)
  8. Van Der Lely Cornelis (204)
  9. Draghetti Fiorenzo      (198)
  10. Lindoff Bengt                (198)
  11. Inoue Hiroshi                  (194)
  12. Kondo Tetsujiro            (193)

Btw, I know the person on the 10th place ? (and soon it is time for # 200 celebration).

One can note that there are significantly lower numbers of granted patents on the EP top-list (top below 400) compared to the US list (top above 4000). The main reason for this is, I believe, that  seeking patents in US is much more common than seeking EP patents (many patents are only filed in specific European countries and hence will not count within the EP patent umbrella) , and the fact that EP patents are a quite recent patent construction  (first patents filed 1978).

Anyhow, it is nice to be on one top-list at least!

Top-10 Inventors, Granted US patents

Who is the top-inventor in terms of number of patents worldwide? Well, it depends on how you count. Is it granted patents in a specific jurisdiction (US, EP, JP,…), or is total number granted patents worldwide? Unfortunately, there is no formal statistics about this, but in the era of Internet and crowd-sourcing there is a nice wiki site, prolific inventors, where determined inventors with more than 200 granted UP patents are listed. Currently the list consists of 138 inventors. I believe the list is not complete especially for the “low numbers” 200-300, and there will probably be more inventors, living and historical, with more than 200 granted US patents. However, for the top 10 the list is probably complete, and in order to qualify on the top-10 list you need …. 1000+ granted US patents!

Top 10 (12) list (May 9th 2017)

Name  / Active year with published patent / #Granted US patents

  1. Shunpei Yamazaki,   1976-2017:             4803
  2. Kia Silverbrook,         1994-2017:             4742
  3. Lowel L. Wood, Jr,    1977-2017:               1494
  4. Roderic A. Hyde,      2001-2017:               1321
  5. Paul Lapstun,           2000-2017:                1282
  6. Gurtej Sandhu,          1991-2017:                1268
  7. Jun Koyama,               1991-2017:                1257
  8. Leonard Forbes,       1991-2017:                1096
  9. Thomas Edison,      1847(b)-1931(d):      1084
  10. Kangguo Cheng,      2004-2017:              1022
  11. Donald E. Weder,     1976-2017:               1000
  12. George Albert Lyon, 1882(b)-1961(d):   993

Quite amazing guys, the top-2 inventors, with over 4700 patents each, 3 times more than no #3 on the list who has “only” 1494 …

As can be seen there are in fact 11 inventors with more than 1000 granted US  patents, and T.A Edison is nowadays only on 9th place. One can also note that 9 out of the 10 top-inventors still are active, and a majority of them have patents within electronics and semiconductor technologies, indicating how important patents are in these areas.

So, I – with only 251 – still have a lot of catching up to do to be at the top of the list…..

Some facts about inventor productivity

There are quite some academic studies about inventing and inventors and patents. An inventor with many inventions/patents is called serial inventor or prolific inventor. In literature, some define prolific inventors with more than 10-20 patents . With a background in mathematical statistics I’m a bit of a nerd when it comes to statistics, so in this post I will try to answer some questions such as; what is the patent productivity distribution for inventors? How many inventors have more than 10 patents? 100 patents?

Well, it turns out that the distribution of number of patents per inventor also quite well follows the Lotka’s  law .

Lotka’s law describes the frequency of publication (patents) by authors (inventors) in any given field (in this case patents and inventors). It states, in the patent and inventor case,  that the number of inventors making x patents in a given period is a fraction of the number making a single patent, following the formula 1/x^a  where a≈2, i.e., an approximate inverse-square law, where the number of inventors filing a certain number of patents is a fixed ratio to the number of authors publishing a single article. Generally, the relation is


So as an example, assume 10000 inventors having filed at least one patent over a certain period of time. Among them, statistically 1/10^2, i.e  1%, 100 inventors,  have made 10 inventions, and statistically 1/100^2=1/10000 = 1 have filed 100 patents.  So among 10000 inventors you can expect to find one inventor that has filed more than 100 patents.

So, now you have learned something new today ?!


Don’t give up!

OK, so you succeeded to open up your own and your colleagues minds. Together you developed a good idea that ended up in an invention, and then you also spend the time needed for preparing a patent application and finally it was filed. Congratulations! ….

One year later, when you almost forgot about the invention, and working full time with another project, a patent engineer sends you a bunch of papers (or nowadays, links to docs) including a search report from the patent office, the search report includes a lot of “X” and “Y” references. This is patent review buzzword meaning

Your invention is neither new nor inventive ☹!

Many people tend to give up here …. But it is now the “fun” starts!

First of all, the examiner is a human that have judged that your invention is neither new nor inventive, due to extensive search of prior art documents. But, maybe he/she has misinterpreted your invention? Maybe, the aim at the patent office is to make a negative decision in order to make the inventor think once again over what can be protected over prior art?

I will say, this is the way you should think;

The examiner is wrong!!

Then you need to read the prior art document (maybe with help of the patent engineer/attorney) and find arguments for that these documents might describe something different and in combination with the prior art documents will end up in something that is completely different compared to your invention. Then you together with the patent persons need to write an answer arguing that the examiner is wrong. In some cases, you might need to reduce the scope of the claims to overcome prior art. Then half a year later you get a new response/search report from the examiner, maybe still negative (then it starts all over again), some other times the report will be positive.

To obtain a granted patent is an iterative process that can take several years! So you need to have perseverance and never give up to early!

I used to say:

Never ever trust an office action, …. especially not the first one!