Canons and Canards: Recreating Prior Art

by Dennis Crouch

In its petition for en banc rehearing Innovation Sciences raises some great old cases.  En banc petition from Innovation Scis., LLC v. Amazon.com, Inc., 842 Fed. Appx. 555 (Fed. Cir. 2021)(unpublished).  In particular, the petition argues that the Federal Circuit’s decision in the case is contrary to both:

  1. Washburn & Moen Mfg. Co. v. Beat ‘Em All Barbed-Wire Co., 143 U.S. 275 (1892) (aka, The Barbed Wire Patent Cases); and
  2. Juicy Whip, Inc. v. Orange Bang, Inc., 292 F.3d 728 (Fed. Cir. 2002) (Note, this is Juicy Whip II.  in Juicy Whip I, the court rejected a contention that the invention lacked utility because of its stated intent of deceiving the public.  Juicy Whip, Inc. v. Orange Bang, Inc., 185 F.3d  364 (Fed.Cir.1999)).

Both of these cases relate to the type of evidence available to invalidate issued patents.

In The Barbed Wire Cases, the patent challengers had introduced oral testimony from 24 eye-witnesses who each testified that they saw a prior art fence on display a county fair in 1858, and including at least one person of skill in the art (a blacksmith).  But, the Supreme Court found the evidence lacking. The court noted the “unsatisfactory character” of oral testimony “arising from the forgetfulness of witnesses, their liability to mistakes, their proneness to recollect things” in a biased way.

The very fact, which courts as well as the public have not failed to recognize, that almost every important patent, from the cotton gin of Whitney to the one under consideration, has been attacked by the testimony of witnesses who imagined they had made similar discoveries long before the patentee had claimed to have invented his device, has tended to throw a certain amount of discredit upon all that class of evidence, and to demand that it be subjected to the closest scrutiny. Thus, the Court showed serious distaste for all oral testimony proffered to invalidate a patent–regardless of the interest of the witnesses.

Id.  The court confirmed this ruling in Eibel Process Co. v. Minnesota & Ontario Paper Co., 261 U.S. 45 (1923), writing that:

[T]here is not a single written record, letter, or specification of prior date to Eibel’s application that discloses any such discovery by any one, or the use of the pitch of the wire to aid the speed of the mechine. The oral evidence on this point falls far short of being enough to overcome the presumption of novelty from the granting of the patent. The temptation to remember in such cases and the ease with which honest witnesses can convince themselves after many years of having had a conception at the basis of a valuable patent, are well known in this branch of law, and have properly led to a rule that evidence to prove prior discovery must be clear and satisfactory.

Id.

The beverage dispenser case, Juicy Whip, follows along this same line, but also focuses on the use of non-prior art as evidence.  In particular, the accused infringer had “recreated” what it claimed was a prior art beverage dispenser and provided oral testimony.  The appellate court rejected the demonstration as “no more reliable than any of the other oral testimony.”  Id.  A patent cannot be invalidated based upon “oral testimony accompanied by an admittedly non-anticipatory visual aid.” Id.  Juicy Whip was decided by the typical pro-patentee Judge Linn and joined by Judge Lourie.  Judge Mayer wrote in dissent that the oral testimony coupled with the model should be sufficient.

The new petition relating to these issues is Innovation Scis., LLC v. Amazon.com, Inc., 842 Fed. Appx. 555 (Fed. Cir. 2021)(unpublished).  The case involves some sort of wireless communications technology useful in home automation.  The closest prior art was a home automation software called HAL2000 created by Tim Shriver back in the early 2000s.  But, the patents don’t just cover software.  Rather, they are directed to various software-hardware combinations and methods of use of those combinations.

In order to show the use of HAL2000 met all the limitations, Amazon had Shriver implement the old code on some old hardware.  Then, Amazon’s expert witness tested the new creation that we might call HAL2019 to show that the prior art met all the claim limitations.  After hearing those results, the jury sided with the defense and found the claims invalid as anticipated. On appeal, the Federal Circuit affirmed. Although it was a non-precedential opinion, the court was somewhat careful to explain that the recreation was appropriate in this case because “substantial evidence supports a finding that the HAL system on which Dr. Johnson based his opinion was representative of prior-art HAL systems.” Innovation Scis. 842 Fed. Appx. 555.

The petition explains the hardware and configuration for prior art HAL systems  were each uniquely installed by customers and that the particular recreated HAL system was designed in 2019 using the asserted claims as a roadmap rather than actually recreating something known to exist.  And, according to petitioner, this recreated system “was the only prior art relied upon at trial.” (with one minor exception). Although nobody testified that the 2019 system as created actually existed before the critical date, folks did testify that the 2019 system “could have” existed. Based upon these recitations of the evidence, the petition then asks two questions:

  1. Whether it is proper to treat a system built in 2019 for purposes of the litigation that included certain prior art components (i.e., the 2019 HAL System) as an anticipatory reference in the absence of proof that 2019 system actually existed prior to the 2006 critical date; and
  2. Whether the suggestion that a prior art system could have existed before the critical date can meet the clear and convincing evidence standard required to prove invalidity.

I see the issues here as fabulously interesting and important. The practical problem for the petitioner though stems from the law that appears willing to allow recreations of the prior art so long as sufficient documentary evidence supports a conclusion that the recreation is representative of the prior art. And, in this case, the challenger provided a variety of evidence to prove that point, including product manuals, articles, and even a video of the system being demonstrated on Oprah. Despite citing these great cases, in the end, the Plaintiff is really arguing that those bits of evidence are insufficient. And, re-arguing the facts is almost always a losing proposition for rehearing.

= = =

Chief Judge Moore authored the Federal Circuit opinion joined by Judges Prost and Hughes.  Donald Jackson and his team at Davidson Berquist represented the patentee at trial, on appeal, and in this petition for rehearing.  Amazon relied on David Hadden and his team from Fenwick throughout the case.  The Federal Circuit’s appellate procedure bars any opposing response unless requested by the court.

33 thoughts on “Canons and Canards: Recreating Prior Art

  1. 8

    As I understand it, the system needed to be built by the customer combining the available software with available hardware. The question was whether any combinations covered by the claims were made.

    “But, the patents don’t just cover software. Rather, they are directed to various software-hardware combinations and methods of use of those combinations.”

    To me, and I may be wrong as this is not my area, the question for anticipation is whether it necessarily follows that the customer would have combined the software with available hardware to make a system that falls within the claims. If it is only possible that that could have happened, anticipation is not established.

    1. 8.1

      “ the question for anticipation is whether it necessarily follows”

      As pointed out in comment 1, this is the well-known concept of inherent anticipation. As pointed out in comment 1 and it’s associated thread, however, what’s a bit odd is that there is no mention of inherency or the relevant precedents in Dennis’ write-up or in the questions presented for cert.

      Instead (as discuscussed in 5.1) the petitioner seems to want to argue that a single prior use of the invention was not demonstrated with “clear and convincing” evidence and, in fact, could not legally have been demonstrated to that level with the type of evidence presented.

      I can think of one good reason for the petitioner to avoid raising the concept of inherent anticipation. Anyone else want to take a stab at it?

      1. 8.1.1

        And as noted in 5.2, your “musing” of 5.1 misses the mark.

        But please entertain us, Malcolm and do (affirmatively) state the supposed reasoning that you only vaguely hint at.

        Does it have anything to do with the Grand Hall experiment?

        Because you have never engaged on the merits of that discussion.

        Does it have to do with the “inherent” difference between three resistors in parallel from three resistors in series [then multiplied by a million or so]?

        Because you have never engaged on the merits of that discussion.

        Does it have to do with the Printed Matter Doctrine — including the Critical Exceptions thereof?

        Because you have never enraged on the merits of that discussion.

        Speak up son, the floor is yours.

          1. 8.1.1.1.1

            I chuckled Chrissy, as I am sure that you have several references buried there, but I am just not sure any of them are pertinent here.

  2. 7

    Off topic, but I am gratified to see no less an authority than Michael Borella acknowledge that—however accurate Judge Rader’s remarks might have been at the time—“[t]he accuracy of th[e PTAB’s ‘death panel’] moniker is debatable, as the latest published trial data for fiscal year 2022 indicate[] that only 34% of all patents undergoing such a trial eventually have at least some claims invalidated.” It has been a while now that the descriptor “death panel” has been a wild exaggeration, but somehow that disparagement persists.

  3. 6

    Love the ‘Juicy Whip I’ v ‘Juicy Whip II’ distinction.

    (it would behoove “marty’s opinion” — that is, inform him of the sense of utility in the patent law terrain — to review that “I” case)

  4. 5

    This passage was key for me: “The petition explains the hardware and configuration for prior art HAL systems were each uniquely installed by customers and that the particular recreated HAL system was designed in 2019 using the asserted claims as a roadmap rather than actually recreating something known to exist. And, according to petitioner, this recreated system “was the only prior art relied upon at trial.” (with one minor exception). Although nobody testified that the 2019 system as created actually existed before the critical date, folks did testify that the 2019 system “could have” existed.”

    Under inherency, the standard is that the system was the natural result of combining the elements, not that it “could have” existed.

    If the patentee can defeat inherency, then obviousness is a challenge IF, as the patentee alleged, the patent was used as a roadmap to recreate the prior art. Hindsight anyone?

    1. 5.1

      Be aware that we are getting an extremely one-sided view of the facts from the losing party in this case. I brought up inherency issue below because of the same passage you mention. But is the legal issue raised at all in the petition? Are the relevant cases cited? Dennis didn’t mention it and the term doesn’t show up in the “questions asked.”

      Looking at the “questions asked”, the petitioner appears to be asking for some kind of bright line rule establishing that certain kinds of evidence are simply insufficient as a matter of law to establish anticipation in a particular (but not rare!) context. Good luck with that! I think this goose is cooked.

    2. 5.2

      “Inherency” (on its own) appears to be a non focus.

      The focal point here is much more on evidence (per se).

      Not seeing Malcolm’s “bright line” either, as the details ‘inherent’ just do not lend themselves to any “bright line” ruling of law.

      It very much matters on the particulars of “what” and the associated “when” OF that “what.”

      To wit (emphasis added):

      Based upon these recitations of the evidence [recitations being MORE ‘evidentiary’ and LESS ‘patent law per se’], the petition then asks two questions:

      Whether it is proper to treat a system built in 2019 for purposes of the litigation that included certain prior art components (i.e., the 2019 HAL System) as an anticipatory reference in the absence of proof that 2019 system actually existed prior to the 2006 critical date; and

      Whether the suggestion that a prior art system [merely]could have existed before the critical date can meet the clear and convincing evidence standard required to prove invalidity.

      Are just not “the mystery” that Malcolm seems to be struggling with.

      1. 5.2.2

        As I understand it, the system needed to be built by the customer combining the available software with available hardware. The question was whether any combinations covered by the claims were made.

        “But, the patents don’t just cover software. Rather, they are directed to various software-hardware combinations and methods of use of those combinations.”

        To me, and I may be wrong as this is not my area, the question for anticipation is whether it necessarily follows that the customer would have combined the software with available hardware to make a system that falls within the claims. If it is only possible that that could have happened, anticipation is not established.

        1. 5.2.2.1

          Absolutely PM- thanks for Adding to the dialogue (in direct contrast to Malcolm’s vapid, inane and complete miss of his typical empty ad hominem.

          Here’s a hint Malcolm: try to at least make an effort to identify my legal point that you do not agree with rather than fumble with such a vacuous and empty try at an insult.

            1. 5.2.2.1.1.1

              You parsed and missed (missed BADLY).

              As is typical for you, you do not actually SAY anything of substance.

              Face it — your mewling of “inherency” misses and my post is simply far more informative.

              As is usual.

  5. 4

    I’d expect that almost every significant software product has a full revision history – the systems used to code software for at least the last two decades can maintain version history line-by-line, forever.

    Does running old code on a new machine “recreate” an invention?

    hmmmm

    Does reading an old book on a new device “republish” the book?

    1. 4.1

      Footnote 1 says “All three asserted patents stem from a common direct continuation application filed on March 15, 2013” which requires really old prior art software, and, as I noted, software software version data and other authentication may not be available from long-gone companies.
      Is this really also an insufficient 102 “inherency” issue? What about the alleged “product manuals, articles, and even a video of the system being demonstrated on Oprah.”? And, just one such clearly dated prior demonstration of the combination of all claim elements would be a “public use.”

      1. 4.1.1

        >And, just one such clearly dated prior demonstration

        Good luck. My general recollection of open source home automation software circa 2012 is that it was directed at high-skill tinkerers (e.g., it was considered normal to roll-your-own PERL scripts). Finding one such nerd who could testify they had a HAL2000 system configured in a specific way (with specific hardware??) on a specific date is going to be hard.

        And that’s before you deal with the legal-side problems about whether one witness’s testimony is enough (vs. 24 in the barbed wire case), whether in-home-only use is “public,” etc.

    2. 4.2

      marty,

      Even taking your post in a colloquial (as opposed to patent law terrain) manner, it is just not clear what point it is that you think that you are trying to make.

    3. 4.3

      > I’d expect that almost every significant software product has a full
      > revision history – the systems used to code software for at least
      > the last two decades can maintain version history line-by-line,
      > forever.

      With respect, your expectation does not comport with reality. Many patents in the software space are litigated a decade or more after their original filing date, a gap of time that is a virtual eternity in that space. Even if the company who made the product relied upon as prior art still exists today (a big “if”), when you try to get the old source code from them, you’ll often find that the source code was revamped several times after the filing date, often accompanied by migration to another version control system altogether. This is because every few years another version control system seems to become more popular or among developers, or as developers leave the company and newer developers join from other companies, they often switch to a different VCS they’re more comfortable with. As a result, even if the company can locate the old (pre-filing date) source code, it often has no way to reliably recreate a previous version control database from a decade or more ago, as anyone who knew how it was configured forgot the details or left the company long ago. I’ve even seen this even with household name software products from Fortune 50 companies.

      1. 4.3.1

        Yes, agreed that version control records may not be available, but if a series of compiled and numbered versions exist at all, its relatively easy to compare the behavior of the differently numbered versions to establish when some feature appeared or was replaced etc.

  6. 3

    If, as it appears, this en banc petition from the Fed. Cir. decision in Innovation Scis., LLC v. Amazon.com, Inc., is attempting to challenge the jury fact-finding, trial and appellate decisions on the basis of insufficient prior art evidence, I find it amazing that the 2011 Sup. Ct. 8-0 decision in Microsoft Corp. v. i4i Limited Partnership is being ignored here? It’s not even listed on the Petitions table of authorities. That much more modern Sup. Ct. decision [even involving software] reaffirmed the old Barbed Wire decisions requirement of “clear and convincing evidence” to invalidate patents.
    [This is not to deny that even i4i may be distinguished here by clearer and more convincing evidence than was presented in i4i]
    The issue here is not trivial – I recall a CPC of Apple [formerly from IBM] years ago saying that the best prior art against many patents in which some specific software feature novelty is claimed is more likely to be found in old obsolete software than in patents. [But if that old software was being regularly edited, proving up its version existing before the patents filing date may be difficult, especially if that software company or its then version date records no longer exists.]

  7. 2

    The development from the Barbed Wire Cases into the modern corroboration rule and its kin is just an overly-aggressive application and reading of that case. It probably won’t be revisited because cases like this one are so rare in our modern days. But it’s a bad line of case-law.

    The way that the Barbed Wire patent case and Eibel have been applied by the Federal Circuit ignore the posture of those cases and the fact that they were bench trials, where there’s more room for the court of appeals to reverse. These days, applying that rule to jury trials creates issues under the Reexamination Clause because it’s fundamentally a witness credibility issue.

    The cases have essentially been applied by the Federal Circuit to require a “two witness” rule for non-documentary prior art (making patent invalidity stand up there with treason). But on top of that, there’s an entire framework used in that line of cases evaluate credibility of witnesses, as a matter of law, based on their relationship with the defendants or alleged inventor of the prior art.

  8. 1

    The way this post is written it seems like the issue is inherency. But it’s confusing as written, I think, because there’s not enough facts. In particular it would be nice to see the claims.

    Also, Dennis, your statement comparing Tim Shriver’s prior art “home automation software” to the claims is very strange. You write:

    “But the patents don’t just cover software. Rather, they are directed to various software-hardware combinations and methods of use of those combinations.”

    Can you explain how Shriver’s software was supposed to perform “home automation” without hardware or without being used?

      1. 1.1.1

        It sounds like he sold software and manuals which (as a matter of necessity) taught implementation on hardware. And then also “there were articles, and even a video of the system being demonstrated on Oprah.”

        Again, the patentee here seems to be making an argument that the challenger improperly relied on inherency to establish anticipation. But it’s not clear exactly what was improper.

        Presumably, given the anticipatory art here, the 103 challenge to the claims was also brutal for the patentee. That would be another reason to just let this deceased dog continue to rest in perpetuity.

        1. 1.1.1.1

          Presumably, given the anticipatory art here

          Why would you jump to that presumption, when the very point at hand is that That cannot be presumed?

          You appear to be indulging (again) in your feelings that the patent holder “must be wrong.”

Comments are closed.