Old-School Submarine Patents

Old-school submarine patents were serious business. Throughout the late 20th Century, more than a handful of patent applicants kept patent applications pending for years-and-years until the claimed invention became market-ready. Once the patent issued, the patent would then be enforceable for seventeen–years from the issue date.  

Two legal changes have eliminated most of the concerning submarine issues for newly filed applications. Patents granted on applications filed after June 7, 1995 have a term of twenty-years from the priority filing date.*  In 2000, the PTO began publishing pending patent applications and now the vast majority of pending applications are published 18–months after filing.

Old patent applications, new patents: I e-mailed with a USPTO official who identified that there are about 600 still-pending patent applications that were filed before the 1995 cut-over (excluding classified applications).  These old cases are all being treated as “special” and the oldest one-hundred cases are receiving special attention.  In 2010, the USPTO issued patents on just over 60 of these old applications. The majority of the 2010 patents were granted to a company known as Personalized Media Communications (PMC). According to its website, PMC still has “over 100 pending patent applications” that were filed prior to the 1995 cut-over.  These patents and applications are all based on a pair of applications filed in 1981 and 1987 but will be in force until at least 2027 — 46 years later.

I looked through the prosecution history of PMC's recently issued patent No. 7,734,251. The final six years of prosecution involved an appeal that was fully briefed (over a 16–month period) and then returned to the examiner for re-briefing. The BPAI eventually decided the case –  affirming the rejection in-part –  then, on re-hearing reversed and agreed with PMC that all of its pending claims were patentable.  The prosecution is also notable for the handful of R.132 affidavits filed by various experts supporting patentability.  In the end, the cases were also being handled at the SPE level. PMC is represented by by Tom Scott at Goodwin Procter.


* The twenty-year term is not triggered by priority claims to either foreign national or provisional patent applications.  Most patent terms are lengthened based on statutory patent-term-adjustment (PTA) due to unreasonable USPTO delays in issuing the patent.

Patent Term: Comparing 17-years-from-issue to 20-years-from-filing

One of the most important attributes of a patent is its term or duration of enforceability.  In 1995, the US patent system began calculating patent term based on the priority filing date of an application rather than a patent's issue date. Under the prior rule, a patent would remain in force for 17 years from the date of issue. Under the “new” system, the term is 20–years from the priority filing date. The three extra years were intended to account for a typical prosecution delay of three years. In addition, Congress added a patent-term-adjustment (PTA) provision to lengthen the patent term in instances where the USPTO unduly delayed in issuing a patent.

Study: For this study, I wanted to consider how the change has actually impacted patent term. To do this, I created a database of 50,000 recently issued patents and calculated the term under both the old and new system.  For the calculation under the new system, I added 20–years to the earliest priority date (excluding provisional and original foreign filing dates) and then added any patent-term-adjustment.  To calculate the term under the old system, I added 17–years to the issue date unless the patent had a family member (e.g., continuation application) with an earlier issue date. In that case, I added 17–years to the issue date of the family member based on an assumption that a terminal disclaimer would have been filed in the case. Because of some potential data discrepancies, I excluded the extreme 1% of results and also patents on applications filed proir to the 1995 change-over. 

Results: Most patents (64%) have a longer term under the new system than they would have been entitled under the older system. Under the new system, the median patent has a term of about 8–months longer than if the term had been calculated under the old system. For 43% of the patent, the terms are are within 1–year of what would have been calculated in the old system; and 89% are within 2–years. Of course, there are some extremes. As you might imagine, patents that benefit most under the new system are those that have the greatest patent term adjustment (PTA).  Patents in TC 1600 (Biochemistry and Organic Chemistry) comparatively fared the worst.  Those patents on average lose 11–months of patent term in the new system as compared to the old system.

Cross-Border Inventors

The number of inventors per patent has been steadily increasing over the past forty years. Patents issued during the past six months, have an average of 2.7 inventors per patent. In all, 68% of these patents list multiple inventors with 13% listing five or more inventors. Prior to 1990, most patents listed only one inventor.

I wanted to look at cross-border or transnational inventing. To do that, I pulled-up all patents that listed inventors from two or more separate nations. For this, I used the inventor's country of residence that is typically supplied when a patent application is filed. Patents with a US inventor (US patents) had an 8% rate of including a cross-border inventor. For US patents, Germany was the most likely country of residence for collaborators. Of the top-10 patenting countries, Japan had the lowest rate of cross-border inventorship at 3% while 48% of Swiss patents listed a non-Swiss inventor. The table below provides more comparative values.

Inventor Country 

Percent that Included Cross-Border Inventor 

Most Likely Country of Co-Inventor 



















Great Britain
























[Updated] Assignees with the most cross-border patents include IBM, Microsoft, Hong Fu Jin Precision Industry (Shenzhen), Intel, Cisco, GE, Schlumberger, Nokia, HP, ExxonMobile, Novartis, and Infineon. Several readers e-mailed and correctly noted that continental European cross-border co-inventorship should be expected because many Europeans cross a border for their jobs. Apparently, Novartis' main building is in Basel Switzerland, but is so close to the border that one Novartis parking lot is actually in France. University of California has a large number of US-China co-invented patents that likely arise from Chinese graduate students seeking their PhD in the US.

Note on the data: The table above includes a classification for "inventor country." The "inventor countries" of a patent include all countries that are designated as the country of residence for at least one of the inventors. Thus, a single patent may be classified as both a US patent and a German patent if there were inventors from each nation. The inventors submit their country of residence as part of the patent application process. The country of residence may be different from the nationality or correspondence address.

Note on spelling: I injured my shoulder this past week and will now be typing with one hand for a while.


USPTO Preliminary Examination Report for FY2010

By Dennis Crouch

Earlier this week, I wrote about the dramatic increase in the number of patents being granted in calendar year 2010 as compared with prior years. [Link] Several outlets expanded upon my report and condemned the USPTO based on a conclusion that the higher issuance numbers must have resulted from reduced examination standards. I disagree with that conclusion. The reality is that we have no evidence of reduced standards. Along with more allowances, FY2010 also saw more abandoned applications, more office action rejections, and more final rejections.

Most patent applications are associated with at least one other family member and multiple claims that can each be amended during the examination process. In that context, it is wrong to think about the patent examination as simply seeking a yes or no answer. In the recent past, however, many patent examiners took their role like that of a professional baseball umpire – calling balls and strikes, i.e., rejecting and allowing. That approach is probably not the most efficient way to reach the more nuanced answers necessary in the patent context. The USPTO's new model is to approach examination in a way that allows for more applicant-examiner dialogue and that that hopefully leads to better patents with fewer rounds of negotiation.

Of course, a major factor in the general increase in cases being concluded may be the reluctance of many corporations to spend money fighting protracted examination battles.

The USPTO has provided some preliminary data looking-back on patent examination during Fiscal Year 2010. The first chart shows the number of patent applications allowed each fiscal year for the past decade. FY2010 saw a 27% rise over FY2009. At the same time, the total number of abandoned applications has increased about 7%.

The second chart (below) counts the number of allowances per examiner (APE). As the chart shows, the APE figure has increased over the past two years. However, the APE rate is much lower than 2001-2004.

Patent Claims Continue to Drop


The chart above shows the percentage of patent applications published with more than twenty claims, grouped by publication date. I use the twenty-claim threshold because patent applicants are required to pay a surcharge to the USPTO for each claim over that limit.

As the chart shows, the percentage of applications with more than twenty claims has been declining for much of the past decade. This change is even more dramatic when looking at applications with forty or more claims.


Of course, if the percentage of high-claim applications is decreasing, the percentage of mid– or low-claim applications must show some increases.  The chart below presents a time series showing the percentage of patent applications with fifteen-to-twenty claims.


Facilitating Freedom-to-Operate Searches

by Dennis Crouch

A good freedom-to-operate (FTO) search will include a search of pending patent applications with an eye toward potential future threats. Over the past decade, the USPTO has published over 2.5 million patent applications. One problem with using that database for freedom-to-operate searches is that most of the published applications are either (1) already patented or (2) abandoned. Over the past three months (June –  August 2010), for instance, about 80,000 published applications were “disposed of.” If we include unpublished applications, during that period, 36,862 US utility patent applications were abandoned and 62,686 utility patents were issued.

Patent academics argue about the vague nature of pending applications (largely because claims can be modified prior to issuance). However, up to now, the database of published applications has been difficult to use as a FTO search database for a far simpler reason — most of the results in a patent application search are for applications that are no longer pending.  Those concluded cases unnecessarily clog the results.

As far as I know, no search-software-vendor has taken on the task of offering US patent application search results that include only pending applications.  Part of the difficulty of accomplishing that result is that it has been difficult to identify which applications have been abandoned.  

As part of its data transparency project, the USPTO recently began publishing a list of abandoned patent applications. [LINK]. In the near future, I expect that this newly available data will be helping to facilitate your FTO searches.

* * * * *




Examiners Ignore Applicant-Submitted Prior Art

Professors Mark Lemley, Chris Cotropia, and Bhaven Sampat recently released a draft of their new article titled “Do Applicant Patent Citations Matter? Implications for the Presumption of Validity.” [Download Here.]

For the article, the trio analyzed the file histories of 1,500+ utility patents issued in 2007 and compared references used in office action rejections with the list of references cited on the patent cover-pages.  The objective was to figure-out the role of applicant-cited prior-art in the examination process.

Findings: Patent examiners rarely rely on applicant-submitted prior-art when making rejections.  Only 13% of the prior art used in office action rejections was applicant-submitted (despite the fact that 74% of cited references are applicant-submitted). Generally, the study found that examiners “effectively ignored” applicant-submitted prior art regardless of how few or how many references were cited; regardless of the timing of the IDS filing; and regardless of whether the submission included an EPO search reports identifying the references as “X-references.”

Implications: The authors suggest several implications of their findings: (1) That it likely does not make sense to find inequitable conduct when an applicant withholds prior art (since the art would not have been used in a rejection anyway); (2) That the presumption of validity associated with patents may be too strong; and (3) That studies based on patent citations likely lack merit.

Notes: I have privately e-mailed Professor Lemley with several comments on the article. Patent prosecutors will not likely be surprised that US examiners tend to rely on their own search results. However, I was surprised at the extent to which they found that this occurs.  As a matter of patent office policy, I would think that some minor changes could alter these results. For instance, examiners should be given a tool for performing text-searches that are limited to submitted (and identified) prior art references.  The office may also want to educate examiners on how to read European search/examination reports.

There are several rational reasons for examiners to cite their own prior art. Because of the backlog, PCTs, and provisional applications, US examination often begins several years after the application was originally filed.  During that interim, many references become available that were not known at filing.  Thus, it is not surprising that applicants rarely cite 102(e) prior art, but examiners cite loads of it.  There is some reason to think that this “newer” prior art is probably better because of technological developments.  It may also be true that the applicant and examiner references are cited for different purposes — namely, applicants cite references that are generally relevant to the invention while examiners are looking for references that teach each particular element in the filed claims.  A third issue is that applicants tend to modify their claims during prosecution. That modification may make their originally cited art less relevant. 

In the words of Lawrence Solum: Download it while its hot.

Revisiting RCE Statistics

In a prior post, I wrote about the rising number of requests for continued examination (RCEs) based on preliminary 2009–2010 numbers.  The charts below provides updated figures through June 2010 and is based on data provided by the USPTO.  The primary difference between these figures and those presented in my prior post is that the prior post under-estimated FY2009 figures.  Although the number of RCE filings continues to rise, this new data indicates that the growth-rate has slowed.

Patent Commissioner Bob Stoll’s recent comments prompted this revisit of the data. I briefly discuss those comments below.


The first chart (above) shows Requests for Continued Examination filings (and earlier forms of CPA and R129 filings) as a percentage of the total number of non-provisional patent applications filed each year.  The denominator here does not include RCE filings but does include continuations and CIP filings. Thus, the calculation is: (Number of RCEs)/(Number of Non-Provisional Filing). The denominator does include plant & reissue applications, but those numbers are negligible.  You may note that my calculations differ from those offered by the USPTO because the USPTO counts an RCE filings as a new application.

The second chart (below) provides raw numbers for the number of non-provisional applications and RCE filings for each year. Again, the non-provisional number does not include RCEs.  A second caveat is that the figures for FY2010 are estimated by extrapolating from current data that goes through June 2010.


The USPTO under Director Kappos and Commissioner Stoll have been focusing some attention on ways to reduce RCE filings. Specifically, the office believes that a good percentage of RCE filings are classifiable as “unnecessary” because they involve issues that should have been resolved beforehand.  In a recent guest post on the Kappos-Blog, Commissioner Stoll pointed to several initiatives that are designed – in part – to reduce the number of unnecessary RCE’s filed:

  • Redesign of the examiner count system that reduces an examiner’s reward when an applicant files an RCE.
  • Redesign of SPE performance plan to focus on “actions per disposal.”
  • Focus on interviews (projected to be up 60% over last year) as a way to understand important issues early-on.
  • Slowing examiner response to RCEs (because of this “RCE inventory has gone from 17,209 as of July 1, 2009, to 35,569 as of July 1, 2010”).

The USPTO appears to continue to be very open to ideas and suggestions that can help the PTO improve their process in ways are sensitive to applicant costs and rights. On that point, I am still working through the 230+ reader comments on my recent query asking Why do Applicants File So Many Requests for Continued Examination?

Updated Statistics: Patent Application Pendency


The above chart shows average application pendency for issued utility patents grouped by issue year. The chart presents three separate series that differ according to the filing-date used in the pendency calculation. The lowest data-series uses a traditional calculation that calculates pendency based on the actual US non-provisional filing date of the application in question without regard to any priority claims. The middle data-series calculates pendency based on the earliest US Priority claim. The highest data-series includes claims to both US and Foreign priority filings.

This chart is useful both for applicants and for third-parties considering moving forward with product development. Importantly, when a patent application is originally filed abroad, it usually take more than six-years before the first US patent is issued.

Requests for Continued Examination Continue to Rise


The chart above shows Requests for Continued Examination (and the earlier CPA/R129 processes) as a percentage of the total number of non-provisional patent applications filed each year. 

Notes: The denominator here does not include RCE filings but does include continuations and CIP filings. Thus, the calculation is: (Number of RCEs)/(Number of Non-Provisional Filing). The denominator does include plant & reissue applications, but those numbers are negligible.


Federal Circuit Extends the Scope of 102(e) “Secret Prior Art”

In an important case, the Federal Circuit has expanded the scope of so-called 102(e) “secret prior art.” Under the decision, a US patent or published application will be considered prior art as of the filing date of its qualifying provisional application.  The case is important because of the large number of provisional patent applications being filed each year.

* * * * *

In re Giacomini (Fed. Cir. 2010)

Giacomini’s patent application was filed on November 29, 2000.  In his search, the examiner found U.S. patent 7,039,683 (the “Tran patent”) and asserted that patent as prior art over Giacomini.  The Tran patent application was filed in December 2000 (after Giacomini) and issued in 2006. However, the USPTO asserted that the patent should be considered 102(e) prior art because it claims priority to a U.S. provisional application that was filed in September 2000.

35 U.S.C. 102(e)(2) bars patentability if

the invention was described in . . . a patent granted on an application for patent by another filed in the United States before the invention by the applicant for patent, except that an international application filed . . . shall have the effects for the purposes of this subsection of an application filed in the United States only if the international application designated the United States and was published under Article 21(2) of [the PCT] in the English language.

The question in this case is whether the 102(e) priority date for prior art reaches-back to the filing date of the provisional application. On appeal, the Federal Circuit agreed with the USPTO that the provisional filing date is the 102(e) priority date.

In the 1968 case of In re Klesper, the Federal Circuit predecessor court (the CCPA) held that the 102(e) follows the prior precedent of treating a prior art disclosure found in an issued patent as being disclosed as of the “filing date of the earliest U.S. application to which the patent is entitled, provided the disclosure was contained in substance in the said earliest application.”  The Federal Circuit agreed that this provision applies equally to provisional patent applications “ so long as the provisional application [provides] written description support for the claimed invention . . . in accordance with Section 119(e).”  Giacomini never argued that the provisional failed to describe the invention found in the prior art.

Therefore, the Tran patent “shall have the same effect,” [Citing 119(e)] including a patent-defeating effect, as to the claimed invention as though it was filed on the date of the Tran provisional. Accordingly, Giacomini, who filed his application after Tran filed his provisional application, cannot receive a patent covering the same subject matter under 35 U.S.C. § 102(e).

This decision follows the BPAI’s 2008 precedential of Ex parte Yamaguchi, 61 U.S.P.Q.2d 1043 (BPAI 2008). However, the decision is in tension with the Hilmer doctrine. In re Hilmer, 359 F.2d 859 (CCPA 1966) (a U.S. application’s 102(e) priority date does not extend to its Section 119 foreign filing date). Giacomini had argued that provisional applications should be treated like foreign filings rather than like non-provisionals because the priority statute for provisional applications is also found in Section 119 and becaues provisional applications lack the formality of non-provisionals. 

Note: US Court interpretations of 103(a)/102(e) offer a major difference between US practice and European practice. Namely, in the US secret 102(e) prior art is available to be combined as part of an argument for obviousness while in most European countries, the filing date of prior art is only important for novelty purposes.

Prep & Pros: Separated by a Backlog

PatentLawPic1010The world continues to turn even though several hundred thousand patent applications are sitting examined at the US Patent Office.  During this multi-year delay, patent attorneys change jobs, receive promotions, raise billing rates, etc.  One bottom line is that the patent professional that drafts a patent application is often not the same individual that argues the prosecution.  This is a subtle problem created by the backlog.

For the following quick survey, please consider the last five USPTO office action responses that you signed.  Out of those five cases, how many of the corresponding US applications did you personally draft & file? (Obviously, this poll is primarily directed to individuals that prosecute patent applications).

[Don't forget to click "Vote" after selecting your response.]

Why do Applicants File So Many Requests for Continued Examination?

Of the patents issued thus far in 2010, more than 25% carry the baggage of a file history with at least one request for continued examination (RCE) and that rate is expected to rise. RCEs are also filed in a significant number of patent applications that are eventually abandoned. 

Patent Office management sees the high RCE rate as a symptom of a breakdown in the prosecution system. In their view, after two rounds of negotiation, the parties should have arrived at final positions — either by identifying appropriate patent-worthy claims or else by determining that no-such patentable subject matter exists. At that point, further negotiation with the same examiner makes little sense in the ordinary case. If the applicant and examiner come to an agreement then the case is concluded, otherwise the applicant appeals to a higher authority.

The Problem of RCEs: Although RCEs do generate USPTO revenue, they are troublesome because they frustrate the USPTO's goal of reducing the large backlog of 730,000+ unexamined patent applications. Furthermore, the current RCE fee of $810 ($405 for small entities) do not cover PTO expenses associated with the ongoing examination. The cost differential is made-up with maintenance and extension fees. In most of the RCE'd cases, the applicant and examiner eventually do reach an agreement without an appeal to the Board — but only after more rounds of discussion. This re-working of applications raises costs for patent applicants, delays patent issuance, prevents work on the backlog of cases, and generally makes all the parties look silly.

PatentlyO063No Solution Yet: Historically, the patent examiners count system provided some incentive for examiners to induce the filing of RCEs. The count system has been modified to reduce that incentive. Despite this disincentive, RCE filings do not appear to have dropped in any significant way.  Another growing RCE disincentive involves patent term adjustment (PTA). The long examination backlog is leading to unprecedented PTAs that can add additional months and years to the term of a patent. Although the filing of an RCE cuts-off the ongoing accumulation of (type-B) PTA, folks are continuing to file RCEs.

Explaining why RCEs are Filed: Although not universal, the broadest independent claims of a patent application are often narrowed during prosecution. It is important to recognize, however, that the narrowing amendments typically involve the inclusion of limitations already found in dependant claims. In the negotiation context, the dependent claims begin to look like pre-set negotiable points. It obviously takes time to reach some mutual agreement between the applicant and the PTO. Perhaps the usual case does need more than two rounds. However, my sense is that the high RCE numbers persist because initial office action rejections regularly fail to directly address the claimed invention and (consequently?) necessary claim amendments are not proffered by the applicant until after the final rejection. At that point, an RCE may be required. Of course, even in the best of times RCEs are sometimes unavoidable — such when prompted by the late discovery of important prior art. The additional time delay of RCEs may be helpful in other cases such as when limitations that distinguish the invention from the prior art are buried in the specification.

The PTO is looking for ways to reduce RCEs and more generally reduce the re-working of applications. To that end, I would like to ask two general questions that will help me create a follow-on survey:

  1. What is your explanation for the large number of RCE filings?
  2. What can the PTO do help prosecution conclude more quickly (on average) and reduce the number of RCE filings? (examiner behavior, PTO procedures, negotiation techniques, patent application format…)
  3. Do some patent attorneys disagree with my notion that the amendment process is a negotiation?

Submitting Positive Decisions to the World Patent Offices

A large number of US patent applications have corresponding applications being examined in various office actions around the world. Although the major Patent Offices generally give no legal deference to the decisions of another Office, it makes sense that the decision of a first-examining Office would at least inform the analysis of the second-examining Office.

On that note, is it a regular practice of patent applicants to highlight the positive actions of other offices (such as the allowance of parallel claims)? Are there certain worldwide offices where this type of submission is more common?

Terminal Disclaimers and PTO: Proposal for a Test Case

Obviousness-type non-statutory double patenting is a judicially created doctrine that operates to prevent patentees from improperly extending a patent's effective term through successive patent grants.

The non-statutory double patenting doctrine was created at a time when patent term was measured from the grant date. At that time, it was easy to "play games" with the patent terms by using a series of continuation applications. Because patent term (for new patents) is now defined by the filing/priority date, most of the potential for games has been eliminated because a family of related applications will typically have the same term. However, a growing issue is patent term adjustment (PTA) that increases the patent term for applications with long prosecution pendency. Because of the increased backlog and changes to PTA calculations, it is likely that later-issued applications in a family will have a longer PTA.

Courts and the PTO have agreed that a patentee can overcome a non-statutory double patenting problem by filing a terminal disclaimer that ties the patent-term of a later-granted patent to the patent term of an earlier-granted patent.

The impact of a terminal disclaimer depends upon what was actually disclaimed. Although applicants can draft their own terminal disclaimer, most folks use the form provided by the USPTO. (SB-0025). The PTO Form disclaims the term of the later-issued patent that extends beyond the earlier-issued patent's term. Regarding PTA, the Form expressly ties the later-issued patent's term to the PTA-adjusted term of the earlier-issued patent.

Although not expressly stated, the Form seems to imply that the PTA of the later-issued patent can be cut-short by the disclaimer. That result, however, is not compelled by the law. In theory (and largely in practice) PTA is based on patent office delay. As such, the equitable basis of the non-statutory double patenting doctrine might not apply to require disclaimer of PTA.

A patent applicant may want to pursue a test-case by modifying the terminal disclaimer form to expressly state that the disclaimer does not cut short the PTA of that later-issued patent. Thus, the term of a later-issued patent with a 400-day PTA and a terminal-disclaimer would limited to the term of the earlier-issued patent plus 400-days.