Correcting Patents During an IPR

by Dennis Crouch

Fitbit, Inc. v. Valencell, Inc. (Fed. Cir. 2020)

In this decision, the Federal Circuit makes two important statements going forward relating to inter partes review proceedings:

  • A joined party has full rights to appeal a PTAB final decision – even as to claims or grounds not raised by the joined party.
  • During an IPR, the PTAB must correct errors in a patent when a certificate of correction would be appropriate and when requested by the parties.  (When you read the case, you’ll note that the court doesn’t quite say this, but almost.)

Valencell’s US8923941 covers methods of generating data relating to blood oxygen level, heart rate, and other physical activity.  The claimed method includes two primary steps: (1) sensing; and (2) processing.  The sensing step requires a single device “attached” to a user that includes a motin sensor (accelerometer) and a pulse-oximeter (photoplethysmography – PPG).  The processing step creates a signal that includes heart rate, respiration rate, and a plurality of physical activity parameters.

This setup is now common with many millions of such devices being sold across the country.  According to its court filings, Valencell repeatedly approached various major players–including Fitbit and Apple–looking for partnership & licensing opportunities. The companies refused and Valencell eventually filed a set of infringement lawsuits back in 2016.  Apple & Fitbit then petitioned for inter partes review.

This case is partically an IPR procedure case and so the following timeline is somewhat important:

  • Apple filed its IPR petition on claims 1-13. The Board partially granted the petition in part — only as to claims 1, 2, and 6-13.
  • Fitbit then filed its IPR petition on claims 1, 2, and 6-13 and moved for joinder with Apple’s case. The Board granted that motion.
  • The Supreme Court then issued SAS — barring the practice of partial institution.  As such, the Board issued a new institution decision adding all of the challenged claims (3-5).
  • The Board then found the originally granted claims to be unpatentable, but found no problem with claims 3-5 (“not unpatentable.”)

Apple did not appeal, but Fitbit has appealed — arguing that claims 3-5 are also invalid.

Joinder and Right of Appeal: The first question in the case is whether Fitbit has a right to appeal regarding claims that it did not challenge in the first place.  Here, the Federal Circuit says yes they do get to appeal — holding that a joined party’s rights apply “to the entirety of the proceedings and includes the right of appeal.” The court here was primarily guided by the right-of-appeal statute that states plainly “Any party to the inter partes review shall have the right to be a party to the appeal.” 35 U.S.C. § 319.  And, the joinder provision indicates that the new entity is “join[ed] as a party.” 35 U.S.C. § 315(c).

On the merits of obviousness: I discussed the general scope of claim 1 above.  Claims 3-5 are all dependent and add a few additional limitations:

  • Claim 3 – the processed data is usable with an application specific interface (abbreviated API)
  • Claim 4 – the “the application” generates statistical relationships between the various parameters. [Note – there is no antecedent basis for “the application”]
  • Claim 5 – adds further definition to the stats developed, including “Bland-Altman plots.”

Do the Analysis: As to claim 3, the parties had offered competing constructions of the term application-specific-interface.  The Board chose the patentee’s narrower construction and ruled the claim not-proven-unpatentable. On appeal, the Federal Circuit ruled that the Board should have completed its analysis and walked through whether the claim was obvious under the narrow construction.

Lack of Antecedent Basis: The limitations of claims 4 & 5 lack antecedent basis. In particular, the claims appear to add limitations to “the application” but that portion of the invention is not previously claimed.   The Board didn’t get into its obviousness analysis because of this technical issue with the claims:

Although we agree that the recitation of the term “the application” in claim 4 lacks antecedent basis in claim 1, we declined to speculate as to the intended meaning of the term.   Although Petitioner and Patent Owner now seem to agree on the nature of the error in claims 4 and 5 [citing briefs], we find that the nature of the error in claims 4 and 5 is subject to reasonable debate in view of the language of claims 1 and 3–5 and/or that the prosecution history does not demonstrate a single interpretation of the claims.

Board Op.

It turns out that during prosecution, the patent applicant cancelled a dependent claim discussed the Application, and the stray reference was not caught by either the attorney nor the examiner.

On appeal, the Federal Circuit holds that the Board should have fixed the error here.

The preferable agency action is to seek to serve the agency’s assignment under the America Invents Act, and to resolve the merits of patentability. Although the Board does not discuss its authority to correct errors, there is foundation for such authority in the America Invents Act, which assured that the Board has authority to amend claims of issued patents. See 35 U.S.C. § 316(d). . . . The concept of error correction is not new to the Agency, which is authorized to issue Certificates of Correction.

In particular, all the parties agreed that the error was inadvertent and could be corrected by the Board.

Although the Board states that the intended meaning of the claims is “subject to reasonable debate,” we perceive no debate. Rather, the parties to this proceeding agree as to the error and its correction. The Board erred in declining to accept the parties’ uniform position and correct the error that claim 4 depend from claim 3. With this correction, the rejection of claims 4 and 5 for absence of antecedent basis for “the application” disappears.

Slip Op. On remand, the Board will correct the error and actually judge the obviousness of these three remaining claims.

13 thoughts on “Correcting Patents During an IPR

  1. 2

    The PTAB remand delay and the problems with previously unchallenged claims 3-5 here were due to the intervening unexpected Sup. Ct. SAS decision, and thus should not occur in later IPRs.
    Could not the requiring of the IPR panel to correct undisputed technical patent defects like claim terminology antecedents also be based on the general duty of the PTO not to knowingly issue invalid patents, as some other cases have noted?

    1. 2.1

      I believe the latter is suggested on the last page of this Fitbit decision by “..and does not comport with the Agency’s assignment to resolve patentability issues.”

      1. 2.1.1

        BTW the subject patent shows a classic “reference dump” of 556 referenced patents and publications [subject to the legal fiction that any examiner has had time to examine many of them?]
        One would think that prior art that should have been considered here before allowance of broad claims would have included prior published common knowledge that U.S. astronauts [and monkey space test subjects] had various such body function sensors?


          [subject to the legal fiction that any examiner has had time to examine many of them?]

          There is no such ‘requirement’ to examine.

          The operative word [instead] is ‘consider.’

          And this falls to the exact same type of ‘consideration’ that ANY reference turned up by the Office itself obtains.


          Why worry about the IDS when prior art should have been found no matter applicant’s disclosure? This appears to be just bad examination. Whatever fantasy IDS regime we could think of probably wouldn’t have made any difference.


          If they “dumped” less than 556 references Apple and Fitbit would have filed IPRs citing prior art “overlooked by the examiner”. Have you not been paying attention to what is happening at the PTAB?

    2. 2.2

      Paul, you’re definitely right, but it runs into a catch-22 type issue. The general duty of the USPTO, as you suggest, is to not knowingly issue invalid patents. They should (an usually do) make every effort to do that but they aren’t always perfect (what human driven endeavor is always perfect?). However, the PTAB doesn’t have the authority to review all possible issues related to patent validity. If a challenged patent has significant 112 problems (rather than 101, 102 or 103 issues), the PTAB will have to pass on it and not resolve the problem.

  2. 1

    Another win for big tech and USPTO joint mission to drive inventors out of business with endless post-issuance validity challenges.

    1. 1.1

      Another chance for Josh to bemoan a greedy applicant’s loss as a loss for “inventors”.

      1. 1.1.1

        Why do you say Dr. LeBoeuf was greedy? I have found him to be a very hard working entrepreneur who earned his patent by obeying all the laws and regulations and paying all the fees required of him.


          How are any of those specific things relevant to the merits of his claims? “Obeying all the laws and regulations” or “paying all the fees” does not address obviousness nor does paying all the fees. I don’t recall those as parts of the analysis in Graham v Deere or KSR . Maybe you could, in your legal expertise, point them out to everyone here. An actually, he didn’t “obey” all of the law all illustrated by the dependency error in the claims. The correction to the claim language is necessary because it does not follow all the laws – specifically 35 USC 112 in this case.

          PS. Have you spent every waking moment with him to know that he is always obeying the laws? or can you just say that in your non-legal opinion (considering you are not a lawyer and didn’t go to law school), during the times you were dealing with him, you did not see anything that you thought was outside the law?

Comments are closed.