Extending Liebel-Flarsheim: “distinctly different” embodiment of claim elements must be enabled

ScreenShot049Automotive Technologies International v. BMW, et al. (Fed. Cir. 2007).

ATI’s patent claims are directed to an automotive side impact crash sensor used to trigger an airbag and include the following means-plus-function limitation:

“means responsive to the motion of said mass upon acceleration of said housing in excess of a predetermined threshold value, for initiating an occupant protection apparatus”

Following ATI’s suggestion, the district court construed the means limitation to include both mechanical and electronic sensor structures. Then, based on this interpretation, the district court ruled the patent invalid for lack of enablement. (summary judgment).  

On appeal, the CAFC affirmed the ruling — finding that the specification did not enable the full scope of the claimed invention.

The specification describes the mechanical sensor in detail with multiple figures. The electronic sensor, however, is only shown in a sparse “conceptual” Figure 11 (see image at right) and described in broad terms as potentially operating “through variety of technologies using, for example, optics, resistance change, capacitance change or magnetic reluctance change.”

Enabling One Embodiment Is Insufficient: Following on the heels of Liebel-Flarsheim v. Medrad, this case reaffirms the principle that the full scope of a claim must be enabled. Enablement of a single embodiment is not sufficient. Thus, the patent’s enablement of one mode of practicing the invention (the mechanical sensor) was not sufficient to satisfy the enablement requirement.

Looking back, Liebel-Flarsheim discussed “full scope” enablement as being “at least reasonable enablement of the scope of the [claim] range.” Here, the CAFC further defines that principle by noting that electronic sensors must be particularly enabled because they are “distinctly different” from the enabled mechanical sensor.

Novel Aspects Must Be Enabled: ATI also argued that Figure 11 and its accompanying text were enabling — especially when coupled with the knowledge of one having ordinary skill in the art. This argument was rejected based on the CAFC’s combined measure of law and fact. As a matter of law, enablement of the novel aspects of an invention must be enabled by the specification even if implementation would have been within the PHOSITA’s skill level. As a matter of fact, the CAFC concluded that sensor mechanism must be a novel aspect based on statements in the specification that side impact sensing is a “new field” and that the only available side-impact switch was a crush sensor.

Notes:

  • Its unclear how the enablement doctrine will work in conjunction with the doctrine of equivalents.
  • The summary judgment was reviewed de novo for clear and convincing evidence of non-enablement sufficient to overcome the patent’s presumption of validity.
  • Disclaimer: My former firm, McDonnell Boehnen Hulbert & Berghoff LLP, handled this appeal on behalf of the plaintiff. Although I have not worked on the case, MBHB took-over the case while I was of counsel last year.

39 thoughts on “Extending Liebel-Flarsheim: “distinctly different” embodiment of claim elements must be enabled

  1. Good point Lysenko. US common law with binding precedent, in these days of perfect instantaneous communication, causes violent swings of the legal certainty pendulum. ROW uses civil law, with no binding precedent. Take the EPO as an example. Show me any significant pendulum swings during the close now to 30 years jurisprudence at the EPO. What price the goal of legal certainty?

  2. People need to mellow out about the latest pronouncements from the Fed Cir and take the long view. No matter what these people have to say today, there’s always hope that this or that panel’s creation will be softened or limited by another panel or by the full court or even SCOTUS (which respects the Federal Circuit about as much as the Federal Circuit respects district courts). Heck, if they keep it up, even legislative relief becomes a possibility.

    But mostly we have time on our side. A patent that we draft today likely isn’t going to be before a judge (if ever) for many years to come, especially with the backlog in various arts, and by then the chances are good that by whatever means, the law will have changed. So relax, people!

  3. Jim posted:

    “However, here it appears to be the patentee who pulled the hoist of their own petard.”

    Technically, the correct usage would be, “Hoisted by their own petard,” which literally means, “blown up by their own bomb.”

    One does not pull the hoist of their own petard, one is hoisted by their own petard.

  4. Jim posted:

    “However, here it appears to be the patentee who pulled the hoist of their own petard.”

    Technically, the correct usage would be, “Hoisted by their own petard,” which literally means, “blown up by their own bomb.”

    One does not pull the hoist of their own petard, one is hoisted by their own petard.

  5. “It seemed to me at first that a more fair result would be to restrict the claim scope to the enabled embodiments to thereby permit a finding of validity but non-infringement.”

    Oh no you didn’t!!!!

  6. patent leather-

    I think your pear (pair?) scenario is distinguishable. The novel aspect of your claim 1 is the addition of a chemical, it seems. You’d have to disclose and enable the chemical and its addition, not necessarily the fruit type. (and, if somehow the fruit type was indeed your novel aspect, then claim 1 is probably invalid if only an apple is disclosed)

    Also, to be analagous to the Liebel-progeny cases, you’d have to argue during litigation that your claim covered a novel aspect that you haven’t disclosed (and enabled). I think this is more the lesson from these cases than “draft picture claims of your example embodiments or die.”

  7. Do I understand the consensus here to be saying the following: claim 1) A method to grow fruit comprising, adding [chemical] to a fruit. claim 2) The method recited in claim 1 wherein the fruit is an apple. Claim 3) the method recited in claim 1 wherein the fruit is a pair. Now assume the spec only describes apple, not pair.

    so claim 3 is invalid for lack of enablement? Why? If pairs are known in the art (like electrical sensors are), is claim 3 really invalid? Isn’t the test for enablement undue experimentation?

  8. “Following ATI’s suggestion, the district court construed the means limitation to include both mechanical and electronic sensor structures. Then, based on this interpretation, the district court ruled the patent invalid for lack of enablement. (summary judgment).”

    I submit that the key to this result is “Following ATI’s suggestion…”

    Usually the scope of a means plus function claim is limited to the enabled embodiments dislosed and enabled in the specification (e.g the old TI v ITC case), so my first reaction was astonishment that a court could (a) first interpret a M+F claim to include a non-enabled embodiment and (b) then declare the entire claim invalid for lack of enablement. It seemed to me at first that a more fair result would be to restrict the claim scope to the enabled embodiments to thereby permit a finding of validity but non-infringement. However, here it appears to be the patentee who pulled the hoist of their own petard.

  9. “Here we have confusion due to use of ‘enable’ when what is meant is ‘written description.’”

    Agreed. We could only wish the CAFC would fess up to such confusion in the opinion, since then a plausible argument *could* be made for the outcome of the case. Sorry to quote Burk and Lemley, but they’ve almost got the 112 written description concept almost exactly right (well except for the last word):

    “Because the unique purpose of the written description requirement is to demonstrate what the inventor had in his *possession* at the time the patent was filed, courts have been understandably reluctant to assume that details missing from the disclosure could be supplied by the prior art knowledge imputed to the PHOSITA.”

    link to btlj.boalt.org

    So to the extent that the applicants did not describe the electronic sensor, that could be used as evidence against them that they did not have “possession” of it and therefore are not entitled to claims encompassing such. (And then the lesson would be the patent drafter must take care not to include embodiments/modifications that aren’t fully described, because the lack of description may be imputed against the applicant.)

    But there is unfortunately a much bigger issue of confusion here:

    The concept of PHOSITA is in 103, and it (or the word “ordinary”) is nowhere to be found in 112. Judges Lourie and Rader are also confused (surprise!), and in their opinion they seem to be concerned both about enablement and to a lesser extent about description. Have a look:

    “The [district] court reasoned that the specification failed to provide sufficient details to teach a person of *ordinary* skill in the art how to make and use an electronic sensor.”

    “That general description, however, fails to provide a structure or description of how a person having *ordinary* skill in the art would make or use an electronic side impact sensor.”

    “The mere boxed figure of the electronic sensor and the few lines of description fail to apprise one of *ordinary* skill how to make and use the electronic sensor.”

    “The specification fails to provide ‘reasonable detail’ sufficient to enable use of electronic side impact sensors.” (Boy is that bogus or what, like I care what kind of sensor my car has when I have a head-on collision. Wonder if I should use it the same way?)

    So I ask, “any person skilled”, “ordinary skill”, what’s the difference? Plenty. Huge! The two hypothetically “skilled” persons in 103 and 112 are not equivalent, and never have been. That’s why a fatally defective (under 112) published application can be a statutory bar against you if you file your enabling/describing CIP more than one year after the publication date.* How can that be, you might ask? It’s because a person with ordinary skill (50% of the skilled people being above him and 50% below him) is often characterized by the courts to be an “innovator” and a “problem solver” (just look at KSR), and might have known to fix a fatal defect (e.g. in a reference for 103 purposes), whereas the person at the bottom of the “any person skilled” continuum might not, since he has generally been held to be incapable of innovation on his own for the purposes of satisfying 112.

    *if you don’t believe me, you can try filing a CIP on a published application thirteen months after the publication date, when the published application is under a 112 1st paragraph enablement or description requirement rejection (please don’t though)

  10. stevenfox:

    All the careful claim drafting upfront doesn’t matter if the inventor hasn’t conceived of (and therefore has no description of) an enabled electrical sensor embodiment.

    This patentee’s problem was not that the claim was written in means-plus-function format. The problem was that the claim covered embodiments that were not enabled. If your sample Claim 1 reads on both the enabled mechanical sensor embodiment and the non-enabled electrical sensor embodiment, this panel (with its Liebel-Flarsheim view) would still have reached the same result.

    Patent practitioners regularly try to prompt inventors to come up with alternative embodiments to broaden the scope of independent claims. Often, this isn’t a problem because specification text and one or two additional drawings can be sufficient. But under the facts of this case, one conceptual drawing with very little text wasn’t enough.

  11. Here we have confusion due to use of “enable” when what is meant is “written description.” The author surmised: “As a matter of law, enablement of the novel aspects of an invention must be enabled by the specification even if implementation would have been within the PHOSITA’s skill level.” Close, but confusing, circular, and ultimately incorrect. A better way to state the point of law would be: “As a matter of law, the novel aspects of an invention must be sufficiently described in the specification under Section 112′s “written description” requirement, even if implementation would have been within the PHOSITA’s skill level.”

  12. “I have not looked at the prior art as of date of the invention, but try this simple language”

    I believe claim 2 would work, but claim 1 might not work – see MPEP 2181 (“lever moving element for moving the lever” and “movable link member for holding the lever…and for releasing the lever” were construed as means-plus-function limitations invoking 35 U.S.C. 112, sixth paragraph since the claimed limitations were described in terms of their function not their mechanical structure). But if you added in claim 1 “a sensor, mechanically coupled to the vehicle chassis and responsive to…” you might be fine.

  13. Try this:

    As the late Honorable Judge Jean Galloway Bissell tried to say in Pennwalt was that 112 P6 simply places a limit on the claims. It has been my experience not to use means plus function unless there is no other choice- but there is always another choice -spend more time on the claims – try drafting claims first and then drafting the specification.

    I have not looked at the prior art as of date of the invention, but try this simple language -

    Claim 1 – ……. a sensor responsive to the motion of said mass upon acceleration of said housing in excess of a predetermined threshold value, for initiating an occupant protection apparatus

    Claim 2- … wherein said sensor is a mechanical sensor.

    Both claims 1 and 2 are supported by the specification. Claim 1 covers both a mechanical and an electrical sensor.

  14. “It likely depends on what alternative language was used in place of the M+F language.”

    That’s correct. I think, if the claim had generically claimed “a sensor” or “a switch” and they argued for a broad claim construction that read on the electrical version, the result would have been the same given this panel.

  15. “If this wasn’t a means claim, then do you think the claim would then be valid? I think so.”

    It likely depends on what alternative language was used in place of the M+F language. What did you have in mind?

  16. “Any news on whether Liebel-Flarsheim is being appealed to the en banc court (or the Supremes)?”

    The Liebel-Flarsheim plaintiff is *not* sympathetic under the facts and, given those facts, the case does not present any important issue. There is no chance an en banc court or the Supremes will take this case. None.

  17. I thought means claims were automatically limited to equivalents of embodiments enabled by the description. Unless no means are disclosed there should be no lack of enablement. This is just extending the ridiculous Liebel decision further.

  18. “Any news on whether Liebel-Flarsheim is being appealed to the en banc court (or the Supremes)?”

    Never.

    This has been another edition of …

  19. Interesting that 2/3 judges on this panel were the same as on the Liebel-Flarsheim panel (Lourie/Rader).

    Any news on whether Liebel-Flarsheim is being appealed to the en banc court (or the Supremes)?

  20. … When the point of novelty is thin, there is a temptation to write less. After all, the invention may reside in a very small aspect. However, in such instances, it may make sense to write *more* to allow you to advocate for a lower level of ordinary skill. With a lower level of skill, you can argue that your beefy specification enabled the claims, but the prior art did not (and thus perhaps avoid a 102 reference). In addition, with a lower level of skill, you can more easily argue that the claims are not obvious. But if (as mentioned above) you need a higher level of skill because you wrote a meager disclosure, you run the risk of the claims being found obvious or not enabled.

  21. Malcolm wrote, “It’s just that your invention is limited to, uh, what you actually invented.”

    I think they actually *did* conceive of this idea, but that it was obvious and therefore it wasn’t a true invention. They needed a level of ordinary skill that was just smart enough to implement the electrical sensor from their meager description, but just dumb enough not to find it obvious from the prior art. Unfortunately for them, they hyped the difficultly of the problem (or, in other words, advocated for a low level of ordinary skill) to avoid obviousness. After making that choice, it was difficult to prove the claims were enabled.

  22. Here, the plaintiffs were in a tough spot. Do they argue for a higher level of ordinary skill (and thus make it easier to prove enablement, but also easier to prove invalidity for obviousness) or do they argue for a lower level of ordinary skill (and thus make the case harder for both enablement and obviousness)?

  23. Applying this opinion to a high percentage of international applications, especially Japanese applications re-filed here without review (they love their means claims) will gut the validity of thousands of existing patents.

    The anti-patent/innovation groundswell/big money lobby effort rises to ever higher levels.

  24. You shouldn’t have to describe what is old, but it is a good idea where you think you will need it. Of course they did that in this case (Figure 11) and the Fed Cir still ruled against them, so…

  25. P.S. my example of transforming a means limitation to a non-means/structural limitation wasn’t very good (but I hope you get the point) because the BPAI can invoke 112, 6th even without the word “means” being present in the claim (e.g. when they infer such intent).

  26. And please note that the “any person skilled in the art to which it pertains” of 35 USC 112 is NOT the same as the PHOSITA in 35 USC 103 (the two skill levels are not equivalent). That is, a reference which is not enabling under 35 USC 112 (e.g. a product brochure) may be prior art under 35 USC 103. Also, a published application with errors that someone with “ordinary” skill would know to correct (103) may be insufficient to teach “any person skilled” in the art how to make and use the same (112).

  27. Would the patentee have fared better, I wonder, by not including FIG. 11 and the corresponding disclosure in his specification at all and letting 112, 6th play its part?

    A general rule has been never rely solely on means claims becuase of their uncertainties (e.g. have a claim that says “an initiating device, responsive to the motion of said mass upon acceleration of said housing in excess of a predetermined threshold value, for initiating an occupant protection apparatus” too)… but do stay under 5/25 if the new rules go into effect.

  28. § 112. Specification

    The specification **shall**
    contain a written description of the invention,
    and

    **[a written description] of** the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable

    any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same,

    and shall set forth the best mode contemplated by the inventor of carrying out his invention.

  29. “‘It is the specification, not the knowledge of one skilled in the art, that must supply the novel aspects of an invention in order to constitute adequate enablement.’ Although the knowledge of one skilled in the art is indeed relevant, the novel aspect of an invention must be enabled in the patent.” [Opinion quotes Genentech 1997]

    That to me isn’t quite the same thing as “As a matter of law, enablement of the novel aspects of an invention must be enabled by the specification even if implementation would have been within the PHOSITA’s skill level.”

    The above quote in the opinion seems to say to me that the specification must supply (ie. have a written description of) the novel aspects, so that these novel aspects can then be enabled (by, e.g. the knowledge of the PHOSITA). Dennis’ formulation is quite different. He is saying the enablement must be by the specification itself, a much tighter proposition.

    Regards, Luke

  30. I have some sympathy with defector’s view. Given the fully enabled mechanical version, it would surprise me if a PHOSITA couldn’t produce an electronic version without too much difficulty. Certainly a subsequently filed application for an electronic one would in all liklihood be bad for obviousness.

  31. From the opinion:

    “‘It is the specification, not the knowledge of one skilled in the art, that must supply the novel aspects of an invention in order to constitute adequate enablement.’ Although the knowledge of one skilled in the art is indeed relevant, the novel aspect of an invention must be enabled in the patent.” [Opinion quotes Genentech 1997]

  32. Like “defector”, I was ‘puzzeled’ by “As a matter of law, enablement of the novel aspects of an invention must be enabled by the specification even if implementation would have been within the PHOSITA’s skill level.”

    But is this actually part of the opinion, or is it Dennis’ gloss?

    I can’t find it in the opinion – albeit at a cursory glance.

    Regards, Luke

  33. It appears that the rub in this case is that the point of novelty in the claims was the means clause referred to in Dennis’ post.

    Clearly, one can’t be entitled to all “means” for detecting side-impact crashes when one has disclosed only one such means in any detail and admitted in the specification that the field is “new”. You can’t have your non-obviousness cake and be enabled to eat it, too.

  34. No, this opinion is dog poo. Sure side impact sensing might — *might* — have been new at the time the application was filed, but accelerometers — i.e., acceleration sensors — are very old and very well known. What I don’t get is the statement that, “As a matter of law, enablement of the novel aspects of an invention must be enabled by the specification even if implementation would have been within the PHOSITA’s skill level.” HUH?!?!? Umm, if it is within PHOSITA’s skill, isn’t it by definition, enabling? CAFC is out to lunch on this one, big time. This is another example of the Federal Circuit being completely clueless as to what public policy demands that a patent application disclose, and that is what’s NEW, not what’s OLD. See, e.g., Evans v. Eaton, 20 U.S. 356 (S.Ct. 1822) “[t]he law does not require of patentees to describe new and old, but merely to distinguish new from old. Otherwise a patent would be more complex and voluminous than a Welsh pedigree.” I know there are countervailing rulings, but they are as flawed as this one. I suspect that either this is somehow a result-oriented decision, or ATI’s lawyers screwed this up and may need to make sure their malpractice insurance is paid up.

  35. To work the new system you must split up “inventions within a master invention” at the beginning of IP plannning. When you write the various applications, chances are you will avoid the pitfalls experienced by ATI because each application will have different focused areas of novelty.

  36. “Ah yes, Liebel, I had nearly forgotten.”

    That would be a mistake, my friend.

    You still get exclusive rights to your invention for 20 years from filing. It’s just that your invention is limited to, uh, what you actually invented. The case law is no longer designed to promote innovation by granting claims of enormous breadth to those who make tiny inroads into crowded fields. Rather than striking it rich by “hitting the megamillions” patent jackpot by stumbling upon the Friendlist USPTO Examiner Ever, inventors may need to find satisfaction with maximum yearly earnings closer to the $500 to $750K range. Or are such meager incomes insufficient to motivate inventors to “do their thing”? Time will tell, my friends. Time will tell.

  37. Ah yes, Liebel, I had nearly forgotten. Combined with KSR, the new RCE and claim number rules, and McKesson, it appears that, yes, all four walls are closing in on me.

Comments are closed.