Search Results for: "means plus function"

Means-Plus-Function: Lack of disclosed structure also fails written description (and nullifies priority claim)

Patent

Uniloc v. Sega (Fed. Cir. 2017) (non-precedential)

Uniloc’s U.S. Patent No. 5,490,216 claims priority to two Australian provisional patent applications.  However, the ‘216 patent also includes some amount of new matter.

The ‘216 patent covers a software registration system that uses a particular algorithm at the game-system to create to create a ‘local’ unique ID. The same algorithm is also run on the registration-server to create a ‘remote’ unique ID.  If the IDs match then the software license is confirmed.

The algorithm is claimed as a “unique ID generating means” and the PTAB identified only one algorithm in the patent specification to serve as the ‘structure’ for the means-plus-function construction.

The algorithm, in this embodiment, combines by addition the serial number 50 with the software product name 64 and customer information 65 and previous user identification 22 to provide registration number 66

 

Since that algorithm was not disclosed in the provisional filings, the PTAB found that the patent at issue was not entitled to the priority claim.  And, without that priority date, intervening prior art invalidated the claims at issue.

On appeal, the patentee argued that the PTAB applied an incorrect standard for determining priority.  Notably, priority is not based upon whether a means-plus-function claim is properly drafted under 35 U.S.C. 112p6.  Rather, under 35 U.S.C. 120, a priority claim is proper  so long as the provisional applications disclose the invention as required by “the first paragraph of section 112.”

According to the patentee, 112p6 demands a narrow ‘myopic’ quest in search for specific structure.  On the other hand, the 112p1 priority quest asks whether the prior filings ‘reasonably convey … possession’  Ariad.

On appeal, the Federal Circuit rejected the patentee’s argument – holding that for means-plus-function claims, the priority analysis does boil down to the search for specific structure — noting that the original disclosure must “describe the invention, with all its claimed limitations.”  Lockwood.  Since nothing in the provisional applications was sufficient to support the “structure” requirement of 112p6, the priority claim also failed.

Although the Federal Circuit does not state this expressly, the holding here is basically that, in addition to being indefinite, a patent claim also fails written description when it includes a means-plus-function limitation without disclosing a specific structural embodiment that performs the claimed function.

= = = = =

This is a case where no sufficient structure (here, algorithm) was disclosed in the priority documents; I wonder what would be the result if one structure was disclosed priority document, but more were added in the non-provisional application.

 

Mettler-Toledo v. B-Tek: Limiting Claim Elements to the Preferred Embodiment

Patent, , ,

By Jason Rantanen

Mettler-Toledo, Inc. v. B-Tek Scales, LLC (Fed. Cir. 2012) Download 11-1173-1200
Panel: Bryson, Moore (author), Reyna

547 patentIn my Introduction to Intellectual Property class this week we discussed the classic tension between (1) reading the claims in light of the specification and (2) not importing limitations from the specification into the claims.  In most instances the court attempts to strike a balance between these two canons (exemplified by Phillips).  When means-plus-function claims are involved, however, different rules apply – as this case illustrates.

At issue in Mettler-Toledo was technology for weighing objects, such as large commercial trucks. The first of Mettler's asserted patents, No. 4,815,547, describes an individual load cell; the second, No. 4,804,052, describes a system and method for correcting weight measurements based on the location of objects on a scale.  Following jury findings that the patents were not infringed and that the '052 patent would have been obvious, the district court denied Mettler's JMOL.  Mettler appealed.

On appeal, Mettler challenged a key construction of the '547 patent claims: that several of the means-plus-function terms require a multiple slope integrating A/D converter, and equivalents thereof, rather than any generic A/D converter.  Mettler contended that by limiting the claims in this way the district court erred by importing the structure of only the preferred embodiment into the claims. 

The CAFC rejected this argument.  "Our case law is clear that a means-plus-function claim limitation is limited to the structures disclosed in the specification and equivalents."  Slip Op. at 7.  Here, the patentee chose to disclose only a single embodiment; thus, the means-plus-function claim was necessarily limited to that embodiment.  "If a patentee chooses to disclose a single embodiment, then any means-plus-function claim limitation will be limited to the single disclosed structure and equivalents thereof."  Id.  Nor was a single reference to an "A/D converter" in the Abstract sufficient to compel a broader structural disclosure because it was not linked to any claimed function.

Obviousness: The CAFC also affirmed the district court's denial of JMOL of nonobviousness of the '052 patent.  Mettler's challenge was based on the argument that a prior art reference did not teach correcting for load position, an element of each claim of the '052.  The CAFC rejected this argument, first noting that the issue of what is disclosed by a prior art reference is a question of fact, then concluding that substantial evidence supported the jury's finding as to the teaching of the 'prior art reference. 

B-Tek's Cross Appeal for Sanctions: During the litigation B-Tek requested documents relating to Mettler's manufacturing capacity in order to rebut Mettler's lost profits damages claim.  Although Mettler stated that it had no such documents, during trial a Mettler employee admitted that certain documents existed (a production schedule and plant diagram).  Mettler further admitted that it provided some of these documents to its damages expert.  The district court denied B-Tek's motion for sanctions on the grounds that the possible existence of these documents had been disclosed during a deposition, these documents were of marginal relevance, and that B-Tek could not show any harm or prejudice given that it prevailed on the issue of infringement. 

On appeal, the CAFC concluded that the district court did not abuse its discretion in declining to sanction Mettler, correctly considering the relevance of the documents and the any harm to B-Tek.

Edited to clarify the first sentence. 

Disclaimer in Prosecution cannot be Recaptured in Litigation

Patent

by Dennis Crouch

Traxcell Technologies, LLC v. Sprint Communications Company, No. 20-1440 (Fed. Cir. 2021); Traxcell Technologies, LLC v. Nokia Solutions and Networks, No 20-1440 (Fed. Cir. 2021)

The patentee primarily these cases on claim construction on two simple terms — based largely on statements made during prosecution to skirt the prior art. 

“Location” – some of the asserted claims take various actions related to the “location of [a] mobile wireless device.”  During prosecution the patentee had argued that its location ability was not limited to “a position in a grid pattern” and did not require a grid pattern overlay.  Rather its location sense was more “adaptable” and “refined.”  The courts found this clear prosecution history disclaimer and so the location term is properly construed to require “not merely a position in a grid pattern.”  This construction excused Nokia from infringement, since the accused Nokia system is arranged in a grid of 50-meter-by-50-meter bins.

“A Computer” – the claims all required “a computer” or “first computer.” The problem was that the accused devices performed the various functions across a set of computers. The district court construed “first computer” and “computer” to mean a single computer that can perform each and every function.  That construction was affirmed on appeal after the Federal Circuit reviewed the claims and specification for supporting evidence.

[I]t would defy the concept of antecedent basis—for the claims to recite “the computer” or “said first computer” being “further” programmed to do a second set of tasks if a different computer were to do those tasks instead.

Slip Op.  This interpretation was also confirmed by the prosecution history statement including a responsive argument submitted to the PTO titled ““Single computer needed in Reed v. additional software needed in Andersson.”  Reed is the Traxcell patent; Andersson was the prior art.  One computer requirement meant no infringement. 

Means For: The case against Sprint partially turned on a means-plus-function limitation: “means for receiving said performance data and corresponding locations from said radio tower and correcting radio frequency signals of said radio tower.”  The patentee pointed to an algorithm disclosed in the specification as the corresponding structural disclosure.  Under Section 112(f), a means-plus-function claim “shall be construed to cover the corresponding structure … in the specification and equivalents thereof.”  It was clear that Sprint did not use the identical algorithm, but the question was whether the “equivalent” penumbra of the claim was broad enough.  In order to determine equivalents under Section 112, the district court applied the function-way-result test generally used in DOE cases.  See Applied Med. Res. Corp. v. U.S. Surgical Corp., 448 F.3d 1324 (Fed. Cir. 2006).  Generally, the function-way-result test ends up with a quite narrow band of equivalents.  Here, the court found that the accused activity not proven to be done “substantially the same way” as that claimed and so was not an equivalent. An aspect of this doctrine that continues to create confusion is whether this “equivalent” evidence is an element of claim construction as suggested by the statute or instead post-construction infringement analysis.

Indefiniteness: Some of the claims were also invalidated as indefinite, even after a certificate of correction. In particular, the those claims required a “means for … suggesting corrective actions . . based upon . . . location” but the specification did not disclose how that might take place. On appeal, the Federal Circuit affirmed. “Although Traxcell demonstrated that the structure makes corrections based on other performance data, it hasn’t shown that any corrections are made using location.”

The Federal Circuit uses a simple if-then shortcut for its indefiniteness analysis of claims that include means-plus-function language. If the specification lacks sufficient structure to support the claimed means; Then the claim is invalid as indefinite.  This approach is probably too rule based. Rather, each time the Federal Circuit should use its lack-of-structure analysis to ask does the claim at issue “particularly point[] out and distinctly claim[] the subject matter which the inventor or a joint inventor regards as the invention.” 35 U.S.C. 112(b).

The four asserted patents U.S. Patent Nos. 8,977,284 (“the ’284 patent”), 9,510,320 (“the ’320 patent”), 9,642,024 (“the ’024 patent”), and 9,549,388. 

Wi-Lan v. Apple: “Clarification” or “reconstruction”?

Patent, ,

By Jason Rantanen

Wi-Lan, Inc v. Apple Inc. (Fed. Cir. 2016) Download Opinion
Panel: Reyna (author), Wallach, Hughes

Although precedential, this case doesn’t really break new patent-law ground.  Instead, it offers a data point in the tricky issue of whether a post-verdict modification to the construction of a claim constitutes a permissible “clarification” or an impermissible “reconstruction.”    The opinion also provides an example of how claim construction can operate at multiple levels, in this case by applying to both the meaning of a given term and the meaning of that term in the broader context of the claim.

Wi-Lan, the owner of RE37,802, a patent relating to wireless data communication, sued Apple and others for infringement based on their manufacture and sale of standards-compliant products. Claim 1 of the ‘802 patent reads:

1. A transceiver for transmitting a first stream of data symbols, the transceiver comprising:

a converter for converting the first stream of data symbols into plural sets of N data symbols each;

first computing means for operating on the plural sets of N data symbols to produce modulated data symbols corresponding to an invertible randomized spreading of the first stream of data symbols; and

means to combine the modulated data symbols for transmission.

(emphasis added).  The critical claim constructions related to “modulated data symbols,” which the district court construed to mean “data symbols that have been spread apart by spreading code” and “first computing means,” which the district court construed to be a means-plus-function element linked to particular structure disclosed in the specification.   Applying these constructions, a jury found that Apple did not infringe and that the ‘802 patent was invalid as anticipated.  Following trial, the district judge denied Wi-Lan’s request for judgment as a matter of law (JMOL) on infringement but granted its request for JMOL of no invalidity based on a modified construction of the structure to which “first computing means” corresponded.

The Interdependent Text and Claim Construction

On appeal, Wi-Lan challenged the district court’s denial of JMOL on infringement (effectively challenging the jury verdict of no infringement).  This issue largely turned on an issue of claim construction: whether the district court’s constructions precluded Apple’s noninfringement defense, a defense premised on the argument that the claims required that data symbols be randomized before combining them, while Apple’s products combined the symbols before randomizing them.

On appeal, the Federal Circuit agreed with Apple.  The text of the claim indicated that the modulated data symbols had to be randomized before combining them.  The “first computing means” produces “modulated data symbols corresponding to an invertible randomized spreading of the first stream of data symbols.'”  The second means then combines “the modulated data symbols for transmission.”  Since “[s]ubsequent uses of the definite articles ‘the’ or ‘said’ in a claim refers back to the same term recited earlier in the claim,” slip op. at 10, quoting Baldwin Graphic Sys., Inc. v. Siebert, Inc., 512 F.3d 1338, 1342 (Fed. Cir. 2008), the term “the modulated symbols” in the second means element referred back to the randomized modulated symbols produced by the first means element, thus requiring randomization-before-combination.  This construction was consistent with the specification.

Wi-Lan’s principal contention was that during its pre-trial construction of  “modulated data symbols,” the district court rejected Apple’s argument that the term requires randomization.   But, the Federal Circuit held, the district court’s construction only involved the unmodified, generic term “modulated data symbols.”   Apple’s argument involved the intersection of the full language in the first means element with the use of “modulated data symbols” in the second means element:  “Even though generic ‘modulated data symbols’ do not have to be randomized, the recited ‘modulated data symbols corresponding to an invertible randomized spreading‘ do have to be randomized.”  Slip Op. at 11.  Through its use of “the,” the  second means element referred back to those randomized data symbols.  Since Apple’s products combined then randomized, there was no direct infringement.  (The Federal Circuit also disposed of Wi-Lan’s argument of infringement under the doctrine of equivalents.)

Post-verdict claim construction

Prior to trial, the district court construed the means-plus-function element “first computing means” as “element 12 of Figures 1 and 4, columns 2:6–10, 2:36–40, 2:58–62, 4:2–12, and 4:35–44, and equivalents thereof.”  Using this construction, the jury found the claims to be anticipated by a prior art reference that used real multipliers to randomize the “modulated data symbols” rather than complex multipliers.  In granting Wi-Lan’s motion for JMOL of no invalidity, however, the district court “determined that, although its construction of computing means ‘does
not specifically provide for a complex multiplier,’ a complex multiplier was nevertheless necessary because ‘expert witnesses from both sides agreed that complex multipliers are part of the structure of the ‘first computing means’ as taught by the ’802 patent.'”  Slip Op. at 8.

On appeal, the Federal Circuit concluded that this was an instance of impermissible post-verdict reconstruction.  “‘[I]t is too late at the JMOL stage to argue for or
adopt a new and more detailed interpretation of the claim language and test the jury verdict by that new and more detailed interpretation.’ Hewlett-Packard Co. v. Mustek Sys., Inc., 340 F.3d 1314, 1321 (Fed. Cir. 2003).”  Slip Op. at 16.  None of the corresponding structure in the court’s pre-trial construction mentioned complex multipliers or referred to the patent figure involving complex multipliers.

While a district judge may “adjust constructions post-trial if the court merely elaborates on a meaning inherent in the previous construction,” Mformation
Techs., Inc. v. Research in Motion Ltd., 764 F.3d 1392, 1397 (Fed. Cir. 2014), this was not such a case.  Such “clarification” is allowed because it only makes “plain . . . what should have been obvious to the jury.”  Slip Op. at 17, quoting Cordis Corp. v. Boston Scientific, 658 F.3d 1347, 1355–57 (Fed. Cir. 2011).  Here, the implicit requirement of a complex randomizers was not obvious to the jury, particularly given conflicting testimony over whether the claims actually did require the use of a complex randomizer.  The result was that the district court did not simply elaborate on the meaning of its prior claim construction in the post-verdict opinion, but altered that construction, something that it could not do.

Means Plus Function: What Structure are you Referring To?

Patent

By Dennis Crouch

University of Pittsburgh v. Varian Medical Systems (Fed. Cir. 2014)

Pitt won its infringement lawsuit against Varian with a $100 million verdict. The patent basically covers a digital image processor that watches surgical patients and provides a warning signal if the patient moves. U.S. Patent No. 5,727,554.

A divided Federal Circuit has affirmed – with the division being over the construction of the claimed “processing means comprising means determining movement of said patient from digital image signals, including movement associated with breathing by said patient.”

The patent statute permits patentees to claim elements according to their function using so-called means-plus-function language. 35 U.S.C. § 112(f). However, the properly interpreted scope of the means-plus-function claim terms is limited to corresponding structures found in the patent specification and their equivalent.

Here, the claim construction debate is on whether the corresponding structure for the processing means is a computer processor programmed to perform either (a) a two-step process or (b) a thirty-two step algorithm. The difficulty in interpretation is that the patent does not affirmatively link the claim language with either structure but instead leaves readers to their own conclusions. On appeal, a two judge majority (Judges O’Malley and Lourie) agreed with the lower court that the proper interpretation was the broader two-step requirement while the dissent (Judge Dyk) would have chosen the more narrow interpretation that would have likely led to a non-infringement determination. To the dissent, the “two-step process” disclosed in the specification only offers more functional language – not structure. The dissent explains his position:

The majority’s identification of structure—the two steps of detecting and tracking fiducials—relies on a single statement in the specification that “[a]s will be discussed fully, the patient motion detector 47 detects and identifies the fiducials 39 and then tracks their movement.” The majority concludes that this disclosure identifies the corresponding structure as a two-step algorithm because the specification “links this algorithm to the function of determining patient movement.” But the statement in the specification does not explain how the fiducials are detected and tracked or how performing those steps accomplishes the function of determining a patient’s movement during radiation treatment.

One rule that would likely flip this decision would be if ambiguity in claim language were construed against the drafter (i.e., the patentee).

= = = = =

The majority did reverse the willfulness finding that almost tripled the damage. On that issue, the court found that the infringer’s non-infringement and invalidity defense were reasonable (or at least not objectively unreasonable).

Functional Claims: Morse, Halliburton & Amgen

Patent

by Dennis Crouch

In the patent context, functional limitations describe inventions in terms of their function or intended use, rather than their specific structure or components. Such claims have been subject to much debate and litigation throughout the history of the US patent system. Notable Supreme Court cases like O’Reilly v. Morse, 56 U.S. (15 How.) 62 (1854) and Halliburton Oil Well Cementing Co. v. Walker, 329 U.S. 1 (1946) significantly impacted patent practice and the balance between functional and structural claim drafting.

The pending Supreme Court case of Amgen Inc. v. Sanofi, No. 21-757 (2023) is another example, with the potential to further shift the landscape regarding functional claim limitations. The title of my essay on the case following oral arguments, “Bye Bye Functional Claims,” hints at my outlook. Although the patentee focused on other issues in its briefing, the Justices repeatedly questioned the permissibility of broad functional claim limitations.

(more…)

The Deere v. Gramm Analysis

Patent

by Dennis Crouch

Deere & Co. v. Gramm (Fed. Cir. 2021).

In 1966, the US Supreme Court decided Graham v. John Deere, and established the framework for adjudging obviousness under the 1952 Patent Act. In the processes Graham’s plow patent was deemed obvious.  In this new case, Deere is looking to invalidate Richard Gramm’s patent. This time though the patentee prevails.

Gramm’s  US Patent No. 6,202,395 covers a sensor and controller for the head-height of a combine harvester.  The case is interesting because of its similarity to the spring-loaded shear in the 1966 case. Here, “a flexible sensor arm that engages the soil and is
dragged across the ground” as the giant tractor rolls along.

The particular novel aspects of the invention here appear to be related to what happens when the Combine is placed in reverse — we don’t want to bob to snap off.  The solution is a spring (“biasing means”) for keeping the arm at an incline but allows for displacement when you reverse.  The spring then uses Hooke’s Law to encourage the sensor arm to return to its original position.

Deere petitioned the USPTO for inter partes review (IPR). Although the Board granted the petition, it eventually sided with the patentee and found that none of the challenged claims had been proven obvious. The case has included some partial institution mix-up following SAS, and the CAFC has previously affirmed the holding with regard to some of the claims, this appeal is related to claims 12-26.

Claimed Biasing Means: That spring discussed above – the claims require a “biasing means for urging said arm to a selected inclined orientation relative to
vertical …”  The parties agree that the claim term is properly interpreted as means-plus-function because “biasing” is a functional limitation rather than structural. Further, the patent document discloses a “coiled spring” that performs the claimed function. Thus – as required by the text of 112p6, the claim limitation is interpreted as requiring a coiled spring “and equivalents thereof” that performs the same function.

In the IPR, this limited view of “biasing means” was helpful to the patentee because it narrowed the scope of prior art that read upon the invention.  On appeal, the Federal Circuit confirmed that the prior art reference (US3611286) did not disclose the biasing means. An image from the prior-art patent is below.  To be clear, the reference does disclose a coiled spring, but it did not operate in the same way — to urge the arm to move toward an “inclined orientation.”

 

The true key controversy of the appeal was really about whether this issue should be reviewed de novo as a claim construction dispute or instead with substantial deference as a factual finding regarding the scope of the prior art.  As usual, the appellant wanted de novo review, but the Federal Circuit refused:

First and foremost, we reject Deere’s attempt to obtain de novo review of a factual finding by reframing it as though it presents a claim construction issue. . . . .

Here, after construing the biasing means limitation, the Board determined that [the prior art] did not teach the first function of the biasing means. That is a factual finding subject to appellate review for substantial evidence. . . . Deere attempts to contort the Board’s opinion by suggesting that the Board actually misconstrued the claim in such a way so as to not read on the teachings of Cleveland. We are not persuaded by that argument, which seeks to blur the clear delineation in the law between the two steps of the invalidity analysis.

Slip Op. Notably, the court does not actually do a very good job of explaining why this particular dispute is not a claim construction dispute except that the PTAB did not see it as claim construction either.

Judgment Affirmed in Favor of Patentee

Eon v. AT&T and the role of “Pure Functional Claiming”

Patent, , , ,

by Dennis Crouch

In a major 2014 decision, the Supreme Court raised the standard of definiteness under 35 U.S.C. 112(b) – now requiring that the scope of patent claims be “reasonably certain” to one of skill in the art. Nautilus v. Biosig. Historically, “reasonably certain” is a high standard and has been linked with the beyond-a-reasonable-doubt standard in criminal law. I.e., claim scope that is reasonably certain may also be seen as having its scope defined beyond a reasonable doubt.  The higher standard can also be contrasted with the prior Federal Circuit that only invalidated ambiguous claims that were both insolubly ambiguous and not amenable to construction.

Despite the dramatic potential of Nautilus, the Federal Circuit has largely muted its impact.

The one area where patent-challengers see continued success is when means-plus-function claims lack appropriate structural support in the underlying patent document.  35 U.S.C. 112(f) allows a patentee to claim a “means” for accomplishing a specified function without reciting the actual structure of the mechanism or material used to accomplish the function.  However, as a rule of construction, the statute indicates that the “means” will be construed to “cover the corresponding structure, material, or acts described in the specification and equivalents thereof.”  The impact is that, although the claim appears to broadly cover a “function” it will be construed to be much more limited.  When an applicant follows this claiming approach, but fails to specify any corresponding structure within the specification then the claim is deemed invalid as indefinite.

In Eon Corp v. AT&T (Fed. Cir. 2015), the Federal Circuit has affirmed that Eon’s asserted patent claims are invalid for failing to specify the structure associated with a purely functional claim element.

Eon’s U.S. Patent No. 5,663,757 covers a data processing station that facilitates instant purchases while watching a television program.  The patent was written back when the wort “means” was still popular among patent drafters.  Here, the district court found that eight separate “means” claims were directed to “complex” computer software such as “causing selected themes to automatically display a second menu.”  Based upon that undisturbed complexity fact-finding, the appellate court found that a structure in the form of the software algorithm should have been disclosed.  Because no software algrithms were disclosed, the software means claims failed and were properly held invalid as indefinite.

Of interest here, the Federal Circuit reasoned that the algorithm is necessary to “avoid pure functional claiming.”  Quoting Aristocrat Techs. Austl. Pty Ltd. v. Int’l Game Tech., 521 F.3d 1328 (Fed. Cir. 2008).

A general purpose computer is flexible—it can do anything it is programmed to do. Therefore, the disclosure of a general purpose computer or a microprocessor as corresponding structure for a software function does nothing to limit the scope of the claim and “avoid pure functional claiming.” Aristocrat. As such, when a patentee invokes means-plus-function claiming to recite a software function, it accedes to the reciprocal obligation of disclosing a sufficient algorithm as corresponding structure.

Although the court here highlights the doctrinal point of “pure” functional claiming — claiming function with no limiting structure — the actual facts are that the disclosed microprocessor and does provide some amount of structure.  As with abstract-idea analysis, it seems here that the question is not so black-and-white, but rather whether some magical threshold has been crossed.

The case also offers some hints to the ongoing debate over subject matter eligibility of computer implemented inventions.  In particular, the court reiterated its prior statements that a “the general purpose computer becomes a special purpose computer when loaded with the special programming.”

Means-Plus-Function: Identification of Function and Corresponding Structure Reviewed De Novo

Patent, Patent Cases 2005

JVW Enterprises v. Interact Accessories (Fed. Cir. 2005).

In a case involving a video game controller, the appeals court reversed-in-part, finding that the lower court had misconstrued the means-plus-function limitation.

De Novo: Like other claim construction, the interpretation of means-plus-function language is reviewed de novo.  More particularly, identification of the function and corresponding structure is reviewed de novo.

In this case, the CAFC found that the district court’s opinion “confuses function with structure.”  According to the appellate panel, “function must be determined before corresponding structure can be identified.” 

Function as claimed: The function of a means-plus-function claim is found in the claim language.  In this case, the district court erred in interpreting unclaimed functions into the claims. 

Structure as described: The corresponding structure of a claimed function must be found in the specification — and must be clearly associated with performance of the function.

The CAFC reversed and remanded based on consequences of the err in claim construction.

Finisar v. DirecTV: Software Means-Plus-Function Claim must be Supported by Particular Structure

Means Plus Function, Patent

Finisar v. DirecTV (Fed. Cir. 2008) [This is Part II of the case discussion. Read Part I]

In a cross-appeal, Finisar asked the CAFC to overturn the lower court’s findings that its means-plus-function claims were invalid for failing to disclose any particular structure to perform the claimed function.  A patent applicant is allowed to claim that her invention covers the “means” for accomplish certain results so long as the application also discloses the “structure” for accomplishing the result.  That structural disclosure must be found somewhere in the specification. Means plus function claims are generally easier to write and are thought by many to be quite broad. Over the past decade, however, courts have increasingly been wary of particularly overbroad claims.

In this case, the claimed means and functions were software related. A number of CAFC cases have required that software means-plus-function claims provide more structural disclosure than just a general purpose computer. Rather, the disclosure must be “the special purpose computer programmed to perform the disclosed algorithm.” WMS Gaming; See also Aristocrat. The CAFC sees its structure requirement as quite minimal and easy to satisfy. Here, however, the complete lack of structure leaves the claims so vague that “one of skill simply cannot perceive the bounds of the invention.”

Notes:

  • One of the Finisar MPF clauses reads as follows:
    • “database editing means, coupled to said one or more computer memory devices, for generating a hierarchically arranged set of indices for referencing data in said information database, including distinct indices for referencing distinct portions thereof, and for embedding said indices in said information database . . .”

Discerning the Purpose and Means of Williamson v. Citrix

Patent

by Dennis Crouch

Functional claim language has long been a mainstay of U.S. patent law. In the historic case of O’Reilly v. Morse, 56 U.S. 62 (1853), the famous inventor of the single-line telegraph (Morse) claimed patent rights to the use of electro-magnetism for transmitting a signal–without limit to any “specific machinery or parts.” Id. The Supreme Court found the claim “too broad and covers too much ground.” Although Morse’s claim directed to a single functionally described element was deemed improper, patent attorneys quickly learned that a combination of functionally-claimed elements could work. An example of this is famously seen in the Wright Brothers early aircraft patents. See US821393.

The Supreme Court pushed-back again on functional claims–perhaps most notably in Halliburton Oil Well Cementing Co. v. Walker, 329 U.S. 1 (1946).  In Halliburton, the Supreme Court invalidated Walker’s means-plus-function claims as indefinite after holding that it was improper to claim an invention’s “most crucial element … in terms of what it will do rather than in terms of its own physical characteristics or its arrangement in the new combination apparatus.”  Id. Halliburton can be read as distinguishing between, on the one hand, claim elements that were already well known in the art, and on the other hand entirely new elements.  Well-known functionality will be understood by those skilled in the art as translating to particular machinery or processes.  And, the existence of prior art means that the patent doctrines of obviousness and anticipation will limit unduly broad scope.  On the other hand, newly invented functionality has the potential of resulting in quite broad scope if divorced from the underlying mechanisms of operation.  The Halliburton decision was not a surprise given that the Supreme Court had already repeatedly criticized functional claims.  A good example of that criticism is seen in the 1938 General Electric decision where the court complained about the “vice” of functional claim limitations:

But the vice of a functional claim exists not only when a claim is ‘wholly’ functional, if that is ever true, but also when the inventor is painstaking when he recites what has already been seen, and then uses conveniently functional language at the exact point of novelty.

Gen. Electric Co. v. Wabash Appliance Corp., 304 U.S. 364, 371 (1938).

Halliburton did not stand for long. Rather, the Patent Act of 1952 was expressly designed to overrule Halliburton and revive the use of means-plus-function claims.  The new law, now codified in 35 U.S.C. § 112(b) reads as follows:

(f) ELEMENT IN CLAIM FOR A COMBINATION.—An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.

Id.  You’ll note that the statute does not suggest any difference between well-known or novel elements in their use of means-plus-function language.  One key feature of Section 112(f) is its particularly narrow interpretive scope.  Means-plus-function elements are not interpreted as extending to the full scope suggested by their broad claim language.  Rather, as the statute recites, the scope is “construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.” Id.  Thus, a claim directed toward a “fastening means” appears quite broad on its face as covering all conceivable ways of fastening, but the term would likely be quite narrowly interpreted if the patent specification is also narrowly drafted.  If the specification disclosed a “nail” as its only example, the court would likely limit “fastening means” to only include nails and their equivalents.  Id. (“structure … described in the specification and equivalents thereof.”).

Invalid as Indefinite: Section 112(f) does not tell a court what to do in cases where the specification lacks disclosure of any “corresponding structure.”  The Federal Circuit’s approach is to hold those claims invalid as indefinite under 35 U.S.C. § 112(b). In re Donaldson Co., 16 F.3d 1189 (Fed.Cir.1994) (en banc).  Indefiniteness was also the hook used by the Supreme Court in Halliburton, then codified within Rev. Stat. § 4888.  I am fairly confident that the Supreme Court would affirm the Donaldson precedent if it ever heard such a case. But, as discussed below, this interpretation leads to some seemingly unjust results.

Structure Plus Function: Because of the narrow construction and indefiniteness concern, patent drafters today largely shy-away from using means-plus-function claim language. Still, functional limitations offer powerful exclusive rights if given their full scope. In the marketplace, consumers  ordinarily care more about whether a machine serves its appointed function rather than how it particularly works. A common approach for many skilled patent attorneys is to take advantage of functional limitations while still endeavoring to avoid being categorized as Section 112(f) means-plus-function limitations. This approach is often quite natural as the English language regularly identifies machines and parts by their function: a brake, a processor, a seat.  These terms are inherently functional but are also tied to known mechanical structures.  I call this approach “structure plus function” and it requires a threshold amount of structure sufficient to ensure that the limitation is not MPF. Structure-plus-function regularly results in broader claims than MPF and claims that that are less likely to be invalidated as indefinite. Again, this is nothing new, and they exact type of language discussed by the Supreme Court in its 1938 GE decision. Still, in recent cases the Federal Circuit has been more aggressively putting claims in the Section 112(f) MPF box.  The key case on point is Williamson v. Citrix Online, LLC, 792 F.3d 1339 (Fed. Cir. 2015) (en banc).

The MPF rules are particularly pernicious in situations where the claim element at issue is already widely available in various forms.  Since it is already widely available, the inventors (and their patent attorneys) typically feel less need to fully describe and specify particular examples of the well-known element.  That inventor “feeling” is generally supported by strong precedent that a thin specification is appropriate for elements of the invention already known to the public. This well-known-element situation is also outside of the point-of-novelty policy concerns raised by Halliburtan.  So, ordinarily, well known elements need only a thin disclosure within the specification.  An off the shelf-component might not receive any structural disclosure at all other than identifying its name and how it connects-in with the rest of the invention.  BUT, that all changes if the claim element is interpreted as MPF.  If its MPF, the claim scope is limited only to what is disclosed (and equivalents), and a specification without structural disclosure will render the MPF claim invalid as indefinite.  Thus, we have two entirely different drafting approaches that will depend greatly on how the claim is interpreted.  And, following Williamson there is no longer a strong MPF interpretive presumption based upon the presence/absence of traditional MPF language such as “means for.”  What that means practically is that the threshold is high-stakes, but unpredictable.

Recently, the Federal Circuit decided two Section 112(f) cases that push back a bit on Williamson:

  • Dyfan, LLC v. Target Corp., — 4th —, 2022 WL 870209 (Fed. Cir. Mar. 24, 2022) (“code”, “application” and “system”); and
  • VDPP LLC, v. Vizio, Inc., 2022 WL 885771 (Fed. Cir. Mar. 25, 2022) (non-precedential) (“storage” and “processor”).

None of the claim elements at issue in the cases used the word “means” as a signal of intent.  Still, the two district courts interpreted the limitations as written in means-plus-function form. In addition, both district courts concluded that the claims were invalid as indefinite since the specifications lacked disclosure of corresponding structure.

The first step in 112(f) analysis is to determine whether a claim element should be classified as a means-plus-function claim.  The most obvious clue is whether the element is written in traditional means-plus-function style–using the word “means” and also including some functional purpose.  An element that does not use “means” is presumptively not a means-plus-function claim.  However, that presumption is not longer a strong presumption. Williamson v. Citrix Online, LLC, 792 F.3d 1339 (Fed. Cir. 2015) (en banc).  And, the presumption can be overcome by a showing that the claim “fails to recite sufficiently definite structure or else recites function without reciting sufficient structure for performing that function.” Id. (internal quotation marks removed).  In addition, “nonce words” are seen as effectively using the word “means.” Id.   Taking all this together, we know that Williamson is a quite important decision. It (1) moved away from a strong presumption against MPF interpretation absent “means” language and (2) established the “nonce word” doctrine. Not surprisingly, the case is the most-cited Federal Circuit decision of 2015.  Still, some members of the court appear to see Williamson as potentially going too far, and that can be seen in the Dyfan and VDPP decisions by Judges Lourie (in both); Dyk, Stoll, Newman, and Taranto.

Claim Construction: The question of whether a claim element is in means-plus-function form is part of claim construction. It is thus the role of the court (rather than a jury) to decide and is ordinarily seen as a question of law with de novo review on appeal.  One exception  to this is comes when a party presents evidence leading to underlying factual  conclusions. Teva Pharmaceuticals USA, Inc. v. Sandoz, Inc., 574 U.S. 318 (2015). Those factual conclusions are given deference on appeal.  Id. In some cases, the factual conclusions may conceivably so intertwine with the legal conclusions that the ultimate claim construction should also be treated as a mixed-question of law and fact and thus given deference on appeal.  (This last sentence should be seen as my speculation on the correct rule rather than black-letter caselaw.)

In most situations, claim construction is purely a question of law and comes-in most commonly when the parties are endeavoring to show the level of understanding of a person of skill in the art (POSA).  For MPF interpretation, this type of expert factual evidence appears to be more common than ordinary claim construction as the parties endeavor to show that a particular claim element would be understood as connoting “structure.” See Zeroclick, LLC v. Apple Inc., 891 F.3d 1003 (Fed. Cir. 2018).

Dyfan’s patents are directed to location-based triggers in mobile phones. Coincidentally, I used my idea of this technology in my internet law course recently in our discussion of privacy law–imagining Ads pop-ups based upon a user’s particular micro-location. In our example here, imagine that as you walk near the vacuum cleaner aisle of Target, your phone receives a “targeted” vacuum cleaner advert or coupon.

The Dyfan claim is directed to a “system” and requires “code” to cause a mobile device to output “visual information” based upon “location-relevant information.”  The claim’s “wherein” clause requires certain timing of the outputs (automatic output of visual information after receiving a communication receipt but without requiring a separate “first message”).  The wherein clause ties itself to the “system” (“wherein the system is configured…”) and does not state if the action is also done by the code. Thus, the courts identified this clause as the “system” element.

The district court found these two claimed elements to lack “structure” and therefore interpreted them as MPF elements.  At step two, district court examined the disclosure documents found no corresponding structure within the specification.  The result then is that the Code and System elements were un-construable and the entire claim invalid as indefinite.

On appeal, the Federal Circuit reversed and remanded–holding that the district court had improperly shifted the burden to the patentee to prove that the elements were not MPF.  This presumption of no-MPF exists because the claims did not use the term “means.” In its decision, the appellate court began in an odd way by stating that overcoming the presumption requires “a preponderance of the evidence.”  I call this ‘odd’ because preponderance of the evidence is the standard used for questions of fact rather than questions of law. As I mentioned, claim construction has been repeatedly deemed a question of law except for limited circumstances.  Here though, there was expert testimony that those skilled in the art would understand “code” to be a particular term of art that encompasses a computer program, and that “off-the-shelf-software” was available to perform the specific function claimed.  On appeal, the Federal Circuit began with a presumption that the code was not-MPF, and that presumption was further solidified by the unrebutted testimony indicating structure.  As part of this, the appellate panel also rejected the notion that the code limitation was “purely functional.”

The system limitation is a bit more complicated — certainly the “system” has structure.  The claim actually requires that the system include a physical building. But, the claim also requires that the system be “configured such that” messages are sent in a particular order and based upon particular triggers.  “A system comprising . . . wherein the system is configured such that . . . .”  Section 112 speaks in terms of an “element”, and the court was faced with the question of whether (1) the system itself is the element, or instead (2) the element is the sub-part of the system that performs the claimed function.  On appeal, the Federal Circuit did not decide this question, but instead held that the claims provide sufficient structure either way.  The court particularly noted that the aforementioned ‘code’ is part of the system and it it is pretty clear that the code would be doing the work here as well.

Although the wherein clause does not expressly refer to the previously recited “code,” it references specific functions that are defined or introduced in the code limitations and thus demonstrates that it is the code that performs the function recited in the wherein clause.

Slip Op. Since these claim elements included sufficient structure, the appellate panel concluded that they were not written in means-plus-function form. And, as a consequence the indefiniteness ruling

VDPP is non-precedential, but reaches the same results. And, although decided only one-day after Dyfan, it favorably cites that decision and particularly the shifting of burdens and evidentiary requirements.  VDPP’s patents relate to technology for blending 2-D video streams so that they appear 3-D. US9699444B2 claim 1 is directed to “an apparatus” having two elements: (1) a storage and (2) a processor.  The storage is “adapted to store one or more frames.”  The processor has a number of  claimed functions, including obtaining and expanding an image; creating a ‘bridge frame’ that is a “non-solid color”; and blending various frames for display.   The patent has a 2001 priority date, and the claims here have shifted substantially to capture market development.

The C.D.Cal. district court determined that VDPP’s claimed  “processor” and “storage” elements were both subject to § 112(f) since they “do not describe how to carry out the recited functions—only that they do” somehow accomplish those functions.  That statement appears almost lifted from Halliburton’s definition of means-plus-function claiming.  In another pithy restatement of its conclusions, the district court also identified each term as a “black box for performance of a function.” On appeal the Federal Circuit reversed–holding that the “processor” and “storage” elements contained sufficient structure.

In the case, the district court “summarily concluded that the limitations [were] subject to § 112(f).”  On appeal, the Federal Circuit found that erroneous because claims without the “means” term are presumed to not be MPF.  This issue here seems to get to the heart of the nonce-word debate. My understanding of ‘nonce words’ is that they are presumed to be MPF triggers.  Thus, the district court’s opinion appears to be based on its conclusion that the processor and storage are “nothing more than
nonce words.”  On appeal, however, the Federal Circuit took a different path. Rather than starting with the nonce work question, the court began with the presumption of no-MPF absent the word “means.”  The court then quoted Dyfan’s statement that overcoming the presumption requires evidence.

Vizio was required to provide at least some evidence that a person of ordinary skill
would not have understood the limitations to ‘recite sufficiently definite structure[s].’

Slip Op. (quoting Dyfan).  The district court had avoided this presumption by identifying the terms as nonce words. That approach was apparently wrong: “The court pointed to no such evidence from Vizio, instead summarily concluding that the limitations are subject to § 112(f). . . That is insufficient to overcome the presumption
against application of § 112(f).” Perhaps the best way to reconcile the developed doctrine here is that a district court should not identify a word as a ‘nonce’ word based upon reasoning alone. Rather, that determination requires some sort of evidence.  In VDPP, the court again referred back to Dyfan and shows district courts how to cite the case:

Vizio’s arguments are particularly unpersuasive in view of our holding in Dyfan. In Dyfan, the district court determined that the limitations “code” and “application” were subject to 112(f).
We reversed the district court’s construction of those terms (among others), explaining that the court did not give effect to the presumption against § 112(f). More specifically, we held that the defendant failed to show “that persons of ordinary skill in the art would not have understood the ‘code’/‘application’ limitations to connote structure in light of the claim as a whole.” That same rationale applies here. As explained above, the court ignored that it was Vizio’s burden to rebut the presumption against § 112(f), and Vizio failed to meet that burden.

Slip Op.  I’ll just note that this type of talk of ‘presumption’ and presumably external ‘evidence’ is radically different from how the court usually discusses claim construction.  Rather, claim construction is a question of law where we are searching for the proper construction.  Typically neither party has a burden of proving a question of law.

In VDPP, the district court also pointed to evidence presented that the storage and processor are known structural forms.  The patentee had fallen into the pernicious trap noted above by explaining in the specification that that “processors” and “storage” are “well-known” in the art.  Here, the court concluded that the specification statement that these are “well known” provides good evidence that the terms were not intended as MPF (without further elaboration or explanation of the reasoning here).

The one problem with the court’s analysis in VDPP is that it implicitly concludes that the “processor” counts as the “element” for MPF analysis.  But, the claim is directed to a particular function, not simply a “processor for processing.” Here, for instance, the claim requires “a processor adapted to: obtain a first image frame from a
first video stream; ….” Although a general processor may be well known, a processor “adapted” as claimed was not well known in the art, and the question that the court needs to ask with regard to structure is whether the claim itself provides the structure for performing the recited function.

As you can see, this area is ripe with historical confusion, statutory bright lines, and shifting precedent.  I expect that the court will continue to develop precedent, but I expect that this pair of cases will begin to tone-down the impact of Williamson while also shifting litigation tactics.

Rembrandt v. AOL: Licensing and Indefiniteness

Indefinite, Patent,

By Jason Rantanen

Rembrandt Data Technologies, LP v. AOL, LLC (Fed. Cir. 2011) Download 10-1002
Panel: Gajarsa (Linn), Linn, and Dyk

In 2008, Rembrandt sued Canon and Hewlett-Packard, among others, for infringement of Patent Nos. 5,251,236 and 5,311,578.  Rembrandt contended that Canon and HP infringed by marketing office products containing modem chips capable of implementing International Telecommunications Union protocol V.32.  The modem chipsets in the accused products are manufactured by a third party, Conexant.

At the trial court level, the district court granted summary judgment in favor of the defendants on the basis that because Conexant was a licensee of the '236 and '578 patent, Rembrandt's rights were exhausted and it could not recover from Canon and HP.  The court further held that claims 3-11 of the '236 patent were invalid for indefiniteness as they improperly mixed method and apparatus elements and that claims 1-11 (all of the asserted claims) were invalid for indefiniteness for failing ot disclose algorithms corresponding to functions set out in the claims.  Rembrandt appealed.

Patent Rights Exhausted Through Prior License
The patents-in-suit were originally owned by AT&T, who granted a license to Rockwell International in the late 1980's.  Under a 1995 side letter, AT&T also granted Rockwell the right to sublicense "to any future divested present business of Rockwell."   That license was subsequently assigned to a reorganized Rockwell ("New Rockwell"), and then to Conexant as part of a spin-off of its modem business.  The patents themselves changed hands several times through a series of divestitures and acquisitions until  finally owned by Rembrandt.

On appeal, Rembrandt argued that patent exhaustion did not apply because Conextant was not a valid licensee under the terms of the agreements.  Looking to the terms of the contact, the CAFC held that contrary to Rembrandt's argument, sublicensing did not require AT&T's consent as long as it occured as part of a divestiture of Rockwell's present business.  Nor did the general rule that "the law does not recognize any right of a nonexclusive licensee to assign the license or to further sublicense" apply because the contract expressly allowed for such sublicenses to be granted.   

The CAFC also applied a pro-licensee interpretation of the scope of the license.  In connection with the sublicense provision, the 1995 Side Letter limited that right "only to the extent applicable to products and services sold by the future divested business prior to its divestiture."  Rembrandt contended that this meant the specific models being sold by Rockwell in 1996, when it divested its semiconductor business to New Rockwell.   The CAFC disagreed.  Looking to the 1988 and 1995 agreements, the court noted that neither refers to specific models of modems, instead specifying product types in general, functional terms.  Thus, because Old Rockwell sold modems prior to the transfer to New Rockwell, and because New Rockwell sold V.34 protocol-compliant modem chipsets prior to the divestiture to Conexant, the court concluded that the sublicensing conditions were satisfied. 

Indefiniteness
The CAFC also addressed three separate indefiniteness issues.  The court first affirmed the district court's grant of summary judgment that claims 3-11 of the '236 patent were indefinite because they contained both apparatus and method claims.  "[R]eciting both an apparatus and a method of using that apparatus renders a claim indefinite under section 112, paragraph 2."  Slip Op. at 15, quoting IPXL Holdings, L.L.C. v. Amazon.com, Inc., 430 F.3d 1377, 1384 (Fed. Cir. 2005).  Rembrandt's only argument in response was to request that the court insert apparatus language into claim 3, from which claims 4-11 depended.  Unsurprisingly, the court declined to do so.  "[T]he correction suggested by Rembrandt is "not minor, obvious, free from reasonable debate or evident from the prosecution history.'"  Slip Op. at 17 (quoting district court opinion).

Rembrandt fared somewhat better on claims 1 and 2.  The district court invalidated these claims (as well as claims 3-11), construing them to contain several means-plus-function elements and holding them indefinite due to the patent's failure to disclose an algorithm able to perform the recited functions. 

The CAFC disagreed that two of these elements were drafted in means-plus-function form.  Although the elements contained the word "means," that created only a presumption that the limitation was drafted in means-plus-function format.  "This presumption can be rebutted if the claim limitation itself recites sufficient structure to perform the claimed function in its entirety."  Slip Op. at 19.  That determination involves a 'one skilled in the art' analysis:  "When determining whether a claim term recites sufficient structure, we examine whether it has an understood meaning in the art."  Id.  

Here, there was no dispute that two of the claim elements, "fractional rate encoding means" and "trellis rate encoding means" recited sufficient structure.  Rembrandt's expert testified that these terms were used in publications and published patents in the early 1990's and were self-descriptive to one of ordinary skill in the art, testimony that went undisputed by the parties.   Thus, these elements recited sufficient structure by themselves to overcome the presumption of treatment under § 112, ¶ 6.

The CAFC did recognize a dispute over two other terms, "buffer means" and "combining means."  Here, the panel appears to have agreed with the district court's interpretation of these elements being in means-plus-function format.  However, the panel could not agree that the specification failed to disclose algorithms for these elements.  Remrandt's expert testified that the figures, text, and table in the '236 patent disclose complete algorithms to a person skilled in the art; Canon argued that they did not.  Thus, the court reversed the district court's grant of summary judgment of invalidity of claims 1 and 2 of the '236 patent.

In re Katz (part 2): Indefiniteness of computer processes

Indefinite, Patent, ,

By Jason Rantanen

In re Katz Interactive Call Processing Patent Litigation (Fed. Cir. 2011)
Panel: Newman, Lourie, Bryson (author)

In addition to the discussion of complex patent litigation management procedures, summarized yesterday, In re Katz is also notable for its analysis of the indefiniteness of computer processing-type means plus function elements.

The patents at issue relate to interactive call processing and conferencing systems.  Ten of the selected claims include "means for processing"-type limitations.   For example, claims 96, 98, and 99 of the asserted '863 patent recite a "means for processing at least certain of said answer data signals."  Applying WMS Gaming, Inc. v. International Game Technology, 184 F.3d 1339 (Fed. Cir. 1999) and Aristocrat Technologies Australia Pty Ltd v International Game Technology, 521 F.3d 1328 (Fed. Cir. 2008), the district court found the claims indefinite because the specifications disclosed only general purpose processors without disclosing any of the algorithms used to perform the claimed functions.   

The Federal Circuit agreed with the district court's analysis of three claims.  Under Federal Circuit precedent, "a computer-implemented means-plus-function term is limited to the corresponding structure disclosed in the specification and equivalents thereof, and the corresponding structure is the algorithm."  Slip Op. at 18, citing Harris Corp. v. Ericsson Inc., 417 F.3d 1241, 1253 (Fed. Cir. 2005).  "[B]y claiming a processor programmed to perform a specialized function without disclosing the internal structure of that processor in the form of an algorithm, Katz’s claims exhibit the 'overbreadth inherent in open-ended functional claims,' … in violation of the limits Congress placed on means-plus-function claims in section 112, paragraph 6."  Slip Op. at 19 (internal citation omitted).  Thus, claims 21 and 33 of the '551 patent and claim 13 of the '065 patent, which claim a “processing means . . . for receiving customer number data entered by a caller and for storing the customer number data . . . and based on a condition coupling an incoming call to the operator terminal, the processing means visually displaying the customer number data” but do not disclose an algorithm that corresponds to the “based on a condition coupling an incoming call to the operator terminal” function are invalid for indefiniteness.

The CAFC reached the opposite result on the remaining seven claims – not on the ground that the patent disclosed the requisite algorithms, but on the basis that no algorithm was necessary because the claim elements may simply claim a general process computer:

[i]n the seven claims identified above, Katz has not claimed a specific function performed by a special purpose computer, but has simply recited the claimed functions of “processing,” “receiving,” and “storing.” Absent a possible narrower construction of the terms “processing,” “receiving,” and “storing,” discussed below, those functions can be achieved by any general purpose computer without special programming.  As such, it was not necessary to disclose more structure than the general purpose processor that performs those functions. Those seven claims do not run afoul of the rule against purely functional claiming, because the functions of “processing,” “receiving,” and “storing” are coextensive with the structure disclosed, i.e., a general purpose processor.

Slip Op. at 20-21. 

The CAFC also reviewed the district court's grant of summary judgment on other issues of indefiniteness, as well as written description, obviousness, claim construction, and noninfringement, largely affirming.  Of these issues, patent litigators may want to take note of the court's comment that when dealing with written description issues it is permissible for the court to construe claims whose meaning has not been placed in dispute, "as claim construction is inherent in any written description analysis."  Slip Op. at 27.  Using this seemingly innocuous statement, the CAFC rejected Katz's argument that it was denied the opportunity to demonstrate specification support or proffer expert testimony on the written description issue because it "should have been clear to Katz that the construction of the claims was important to the written description analysis." 

Functional Claim Language in Issued Patents

Patent

By Dennis Crouch

In a prior post, I included a chart showing a recent rise in the use of the functional claim phrase “configured to.” See Crouch, What Does it Mean for a Device to be “Configured To” Perform a Particular Function, Patently-O (2014). Here, I have updated the data to also add information regarding claims that include the term “means” (the traditional primary trigger for §112(f) analysis) and those that include the phrase “for …ing” (such as “for calculating”) without reciting the term “means.” I see these chosen categories as indicia of the existence of functional claim elements. However, I recognize that my rigid automated approach is both somewhat under and over inclusive. However, the results (below) are so dramatic that it is difficult to argue that nothing is going on here. Complicating this, a legal definition of functional claim language (outside the scope of §112(6)) is difficult to nail down. Working a bit on this definition, the Supreme Court in 1946 wrote that a “claim which describes the most crucial element in a ‘new’ combination in terms of what it will do, rather than in terms of its own physical characteristics or its arrangement in the new combination, is invalid as a violation of [the indefiniteness requirement].” Halliburton v. Walker, 329 U.S. 1 (1946).

For the analysis here, I first downloaded the text of all U.S. utility patents issued between 1976 and 2014. I then wrote a script that parses through the independent claims of each patent to determine whether any of the aforementioned phrases are found within those claims. Because the worry with functional claim elements tend to be their over breadth, I looked only at independent claims rather than the naturally more narrow dependent claims. With those results in-hand, I then calculated the frequency-of-appearance in a week-by-week time series. The chart is slightly smoothed by reporting a 10-week moving average – that smoothing largely imperceptible because the time series encompasses almost 2,000 weeks of data). A nice element of using the entire population of utility patents is that there is no need to include significance levels or confidence intervals.

The basic results are apparent from the chart below. The use of traditional Means-Plus-Function (MPF) language in patent claims has dropped dramatically since the early 1990’s and continues to drop. (Blue data series). In the 1980’s most patents included an independent claim written in MPF form. Today, that frequency is less than 10%. The decline in the use of MPF claim language corresponds roughly with a set of cases that cabined-in the construed scope of MPF elements and also bolstered the doctrine that an MPF claim is invalid if the patent document lacks an affirmative statement showing a structure that performs the claimed function. See In re Donaldson Co., 16 F.3d 1189 (Fed. Cir. 1994) (en banc). Although patentees have eschewed the statutory sanctioned MPF format, they have not given up the use of functional claim limitations. In the chart below, I pick up on two particular common approach. One approach has been to replace the term “means for” with a more structural noun such as “circuit for” or “instructions for.” Those types of claim elements are shown in the green data series below and show a rise in that claim format in the 1990s that matches the decline in MPF usage. The red data series shows the use of “configured to/for” language in independent claims as discussed in the prior essay.

As patentees moved away from the use of means-plus-function claim language, they could have conceivably chosen either to focus more particularly on structural elements or instead to continue with an approach of claiming according to the function of the unspecified structure. The chart above graphically suggests that patentees have chosen the latter – replacing means-plus-function claim elements with functional claim elements carefully designed to avoid the statutory trap of §112(6).

The sheer number of patents at issue here is impressive. To be clear, we are talking about millions of patents that are caught-up in this transition. Thus, if we think that it’s a problem to have functional claim limits written outside of the safeguards of §112(6), then what we have here is a big problem.

BPAI Remanding Cases Involving Computer-Oriented Means-Plus-Function Claims

Patent

In Ex party Rodriguez, the BPAI rejected a patent applicant’s means-plus-function (MPF) claims as indefinite for failing to provide any corresponding structures in the specification beyond a general purpose computer. That decision followed the Federal Circuit’s Aristocrat holding that “simply disclosing a computer as the structure designated to perform a particular function does not limit the scope of the claim to ‘the corresponding structure, material, or acts’ that perform the function, as required by section 112 paragraph 6.”

A patent examiner recently wrote-in to identify the fact that the BPAI has recently remanded dozens of pending appeals in light of Rodriguez. The remands use the following form language:

Claims ___ of the instant application contain functional language that may be indefinite under 35 U.S.C. § 112, second paragraph. See Aristocrat Techs. Austl. Pty Ltd. v. Int’l Game Tech., 521 F.3d 1328 (Fed Cir. 2008); Ex parte Rodriquez, 92 USPQ2d 1395 (BPAI 2009). On September 2, 2008, the Deputy Commissioner for Patent Examination Policy, John J. Love, issued a memorandum entitled “Rejections under 35 U.S.C. 112, second paragraph, when examining means (or step) plus function claim limitations under 35 U.S.C. 112, sixth paragraph” . . . Thus, there is a question as to whether claims ___, and the claims which depend upon these claims, meet the requirements of being definite under 35 U.S.C. § 112, second paragraph. . . . Accordingly, it is ORDERED that the application is remanded to the Examiner to determine if claims ___, and the claims which depend upon these claims, meet the requirements of 35 U.S.C. § 112.

See, Ex Parte Chefalas et al, Appeal No. 2009012658, Ex Parte Yamanaka et al, Appeal No. 2009010948, Ex Parte Gassen et al, Appeal No. 2009008147, Ex Parte McCabe, Appeal No. 2009013024, Ex Parte Selep et al, Appeal No. 2009008213, Ex Parte Martin et al, Appeal No. 2009011586, Ex Parte Tosey, Appeal No. 2009007299, Ex Parte Van Doorn, Appeal No. 2009011085, Ex Parte Sohraby et al, Appeal No. 2009012154, Ex Parte Breese et al, Appeal No. 2009008013, Ex Parte Choe et al, Appeal No. 2009010957, Ex Parte Swart et al, Appeal No. 2009011542, Ex Parte Kakinuma et al, Appeal No. 2009007905, Ex Parte Zriny et al, Appeal No. 2010001851, Ex Parte Park, Appeal No. 2010002356, Ex Parte Cherkasova et al, Appeal No. 2010000365, Ex Parte Messick, Appeal No. 2009014361, Ex Parte Levy et al, Appeal No. 2010000512, Ex Parte Gusler et al, Appeal No. 2010000556, Ex Parte Betge-Brezetz et al, Appeal No. 2010001913.

In the Chefalas case, the independent claim in question reads as follows:

48. A data processing system for handling a virus, the data processing system comprising:

receiving means for receiving a notification of a presence of a virus on a client data processing system through a communications link;

severing means for severing communication with the client data processing system through the communications link in response to receiving the notification; and

executing means for executing virus removal processes on the server data processing system.

What does it mean for a device to be “configured to” perform a particular function?

Patent,

By Dennis Crouch Chief Judge Rader recently argued that no patentable weight should be given to claim terms focusing on how a device is “configured to” perform a particular task. See Superior Industries v. Masaba (Fed. Cir. 2014) (concurring opinion). The Chief writes:

A system claim generally covers what the system is, not what the system does. Hewlett–Packard Co. v. Bausch & Lomb Inc., 909 F.2d 1464 (Fed.Cir.1990). Thus, it is usually improper to construe non-functional claim terms in system claims in a way that makes infringement or validity turn on their function. Paragon Solutions, LLC v. Timex Corp., 566 F.3d 1075 (Fed. Cir. 2009).

In Superior Industries, the patentee had claimed a dump truck with a “support frame . . . configured to support an end of an earthen ramp.” On remand, the district court will need to determine again whether that language should be interpreted as a positive claim limitation.

One problem with Judge Rader’s opinion is that other judges do not uniformly agree with his perspective. The diversity of court opinion compounds an already difficult process of understanding what potential structures are covered by function-focused claim terms. A third compounding difficulty is the rapid rise in the use of functional claim terms.

The Patent Act expressly provides for the use of functional claim language — written in means-plus-function format. 35 U.S.C. §112(f). The statutory requirements of §112(f) began being more strictly enforced in the mid-1990s and, as a consequence, means-plus-function limitations have fallen from favor. Meanwhile, patentees have caught-on to using “configured to” language as a rough equivalent of a means-plus-function claim term but carefully drafted in order to avoid actually falling under the ambit of that provision. The benefit then for patentees is that the functional language is broad and not limited to particular structural embodiments.

The chart below shows the percentage of issued patents that include at least one “configured to” claim limitation within an independent claim. This dramatic transformation of claiming strategy raises the importance of cases such as Nautilus v. Biosig Instruments (pending Supreme Court case on the proper standard for indefiniteness) and Lighting Ballast (pending en banc decision on claim construction deference with a focus on functional claim terms should be interpreted as falling within 112(f)).

In Ex Parte Miyazaki, 89 USPQ2d 1207 (B.P.A.I. 2008), the Board of Patent Appeals revived the old precedent of Halliburton Oil Well Cementing Co. v. Walker, 329 U.S. 1 (1946) in holding that purely functional claim language is not allowable unless done within the scope of 112(f).

This general prohibition against the use of “purely functional claim language” (and the more specific Halliburton rule) has not been completely eliminated. Rather, “purely functional claim language” is now permissible but only under the conditions of 35 U.S.C. § 112, sixth paragraph, i.e., if its scope is limited to the corresponding structure, material, or act disclosed in the specification and equivalents thereof.

In Miyazaki, the problematic claim language was “a sheet feeding area operable to feed at least one roll of paper, at least one sheet of paper and at least one stiff carton toward a printing unit.” In Halliburton, the problematic claim language was “a tuned acoustical means which performs the functions of a sound filter.” In each case, the claims were found indefinite.

Is the Federal Circuit too Trigger-Happy Invalidating Means Claims?

Patent

by Dennis Crouch

Rain Computing has filed an interesting petition to the Supreme Court focusing on indefiniteness and means-plus-function claims. Rain Computing, Inc. v. Samsung Electronics America, Inc., No. 21-170 (Supreme Court 2021).   Rain’s asserted claims require a “user identification module configured to control access of … software application packages.”  The Federal Circuit interpreted “module” as a non-structural nonce word.  As such, the court interpreted the module element a means-plus-function limitation under 35 U.S.C. § 112(f).  The problem though was that the specification did not disclose any corresponding structure “clearly linked” with the claimed function.  As such, the Federal Circuit found the claims invalid as indefinite.

The first question in the petition focuses on the courts trigger-happy approach to invalidating means claims and the interpretive difference between a claim limitation and the claim as a whole.  Although courts often construe individual claim limitations, indefiniteness is generally determined by considering the claim as a whole in light of the specification and the prosecution history and asking whether the claim informs “with reasonable certainty those skilled in the art about the scope of the claim.”  Nautilus.  The idea here is that the Federal Circuit should have asked the broader question — would someone of skill in the art have reasonable certainty as to the claim scope?  Instead, the court did a shortcut and effectively held that the  element-by-element corresponding-structure is a requirement of patentability if means language is used. The petitions also notes the construction provision of 112(f) requiring that “such claim shall be construed to cover the corresponding structure …”  On its face, the provision focuses on construction of the whole claim, not just the means-element.

The second question is directed toward the Federal Circuit’s conclusion that the specification disclosed only a ‘general purpose computer’ when the district court held otherwise.  The petition suggests that change-of-course would require clear-and-convincing evidence before the claim can be invalidated.  The underlying assumption of this argument is that some aspect of interpreting the specification is a question of fact (rather than a question of law).

Whether a patent claim may be invalidated for indefiniteness only if, under 35 U.S.C. § 112, para. 2 (now subsection 112(b)) as construed by Nautilus, Inc. v. Biosig Instruments, Inc., 572 U.S. 898, 901 (2014), the whole claim, “read in light of the specification delineating the patent, and the prosecution history, fail[s] to inform, with reasonable certainty those skilled in the art about the scope of the claim” and 35 U.S.C. § 112(f) (previously 112 para. 6) affects only coverage of a combination claim of which an element is expressed as allowed by the subsection?

Whether the United States Court of Appeals for the Federal Circuit panel in this case erred by invalidating a patent as indefinite without mentioning 35 U.S.C. § 112, para. 2 (now subsection 112(b)) and applying the Nautilus test and by substituting, without clear and convincing evidence, its conclusion of a “general purpose computer” for the district court’s finding of no “general computer”?

[Petition for Writ of Certiorari]

= = = =

I’ll note here that Rain Computing is represented by the Hsuanyeh Law Group, founded by Hsuanyeh Chang.  Dr. Chang is also the listed inventor of the patent here and is CEO of the petitioner Rain Computing.

Ring & Pinion v. ARB: No Foreseeability Limitation on DOE

Patent,

Ring & Pinion Service Inc. v. ARB Corporation Ltd. (Fed. Cir. 2014) Opinion
Panel: Moore (author), Clevenger, Reyna

One of the main justifications for the doctrine of equivalents is that it prevents patents from becoming worthless as a result of unforeseeable changes in technology.  As Judge Rader observed in his concurrence in Festo:

A primary justification for the doctrine of equivalents is to accommodate after-arising technology. Without a doctrine of equivalents, any claim drafted in current technological terms could be easily circumvented after the advent of an advance in technology. A claim using the terms “anode” and “cathode” from tube technology would lack the “collectors” and “emitters” of transistor technology that emerged in 1948. Thus, without a doctrine of equivalents, infringers in 1949 would have unfettered license to appropriate all patented technology using the out-dated terms “cathode” and “anode”. 

This case deals with the inverse: whether the doctrine of equivalents can apply when a technology was foreseeable at the time of patenting.

R&P sought a declaratory judgment that its product did not infringe ARB’s patent.  The parties agreed that R&P’s product literally met all of the limitations of claim 1 of the patent except for one, which they agreed was present as an equivalent in R&P’s product.  However, they disagreed over whether the doctrine of equivalents could apply since the equivalent would have been foreseeable to a person having ordinary skill in the art at the time the application was filed.  The parties thus stipulated that infringement in this case turned on the legal question of whether “foreseeability of an equivalent at the time of application prevents use of the doctrine of equivalents.”  The district court ruled that foreseeability did not preclude application of the doctrine of equivalents, but granted summary judgment to R&P on a vitiation theory.

On appeal, the Federal Circuit agreed with the district court that forseeability is not a limitation on the DOE:

There is not, nor has there ever been, a foreseeability limitation on the application of the doctrine of equivalents. It has long been clear that known interchangeability weighs in favor of finding infringement under the doctrine of equivalents.  [the CAFC cites a number of cases discussing interchangeability, including Warner-Jenkinson and Graver Tank]. Excluding equivalents that were foreseeable at the time of patenting would directly conflict with these holdings that “known interchangeability” supports infringement under the doctrine of equivalents. We conclude that the foreseeability of an equivalent at the time of patenting is not a bar to a finding of infringement under the doctrine of equivalents.

Slip Op. at 4-5.

Nor did it matter that the claim element involved was drafted in means-plus-function terms.  The doctrine of equivalents applies equally to these types of claim terms, and there is no “partial” foreseeability limitation here.  “Nothing in Chiuminatta or in any other case cited by R&P supports its assertion that there exists a foreseeability exception to the doctrine of equivalents that applies to means-plus-function or any other claim terms.”  Slip Op. at 7.

The defendant here thus lost on infringement because there is no foreseeability limitation on the use of the doctrine of equivalents.  Furthermore, the stipulation that the parties entered into precluded consideration of other potential issues: the district court erred by not enforcing that stipulation.

Policy rationale: Given that a primary rationale for the doctrine of equivalents is to address the problem of after-arising technology, what is the basis for concluding that the DOE also extends to foreseeable technologies, i.e.: those that the patent drafter could have included in the patent.  While Ring & Pinion v. ARB does not offer an explanation, the Supreme Court’s opinion in Festo does:

“The language in the patent claims may not capture every nuance of the invention or describe with complete precision the range of its novelty. If patents were always interpreted by their literal terms, their value would be greatly diminished. Unimportant and insubstantial substitutes for certain elements could defeat the patent, and its value to inventors could be destroyed by simple acts of copying. For this reason, the clearest rule of patent interpretation, literalism, may conserve judicial resources but is not necessarily the most efficient rule. The scope of a patent is not limited to its literal terms but instead embraces all equivalents to the claims described.”
Festo Corp. v. Shoketsu Kinzoku Kogyo Kabushiki Co., Ltd., 535 U.S. 722, 731-32 (2002).

In other words, patent drafters are not omniscient; they are human beings who are using an imperfect tool (language) to describe an intangible concept.  In addition, patent drafters are only one person or a few people, whereas competitors seeking to design around the patent may be numerous (the game show 1 vs. 100 was premised on a concept like this).  Under this view, the doctrine of equivalents tries to level the playing field.

Furthermore, although not discussed by the court, foreseeability remains an important component of the prosecution history estoppel analysis.  Under Festo, the presumption arising from a narrowing amendment can be rebutted if the patentee can show that the equivalent was unforeseeable at the time of amendment.  535 U.S. at 741.  But Ring & Pinion did not assert that there was a narrowing amendment, so this application of foreseeability was irrelevant.

 

 

Standing to Defend your (Expired) Patent

Patent

Sony Corp. v. Iancu (Fed. Cir. 2019)

Sony’s patent no. 6,097,676 is directed toward multiplex audio recordings with each channel having a different language translation.  Back in 2015, SONY sued ARRIS for infringing the ‘676 patent. (One of several infringement lawsuits between the companies).  In 2016, ARRIS turned-around and filed for inter-partes-review (IPR). The PTAB instituted as to two claims (5 and 8) and found them unpatentable (obvious).  That final written decision was issued in September 2017. ARRIS and SONY subsequently settled their ongoing litigation and in November 2017.  At that point SONY appealed and the USPTO intervened to defend the PTAB decision.  The patent had also already expired (August 2017).

On appeal, the majority and dissent argued over standing — whether there was any ongoing case-or-controversy since the patent expired in 2017 and the parties settled. Judge Newman – The Great Dissenter – argued that the court lacked jurisdiction:

There is no interest, neither private interest nor public interest, in the fate of this patent. There appears to be no consequence of either our appellate decision today or the potential PTAB decision on the remand now ordered by the court.

The majority opinion authored by Judge Dyk and joined by Chief Judge Prost held otherwise:

It is well-established that our decision (and the Board’s decision on remand) would have a consequence on any infringement that occurred during the life of the ’676 patent.

I agree with the majority here. The patentee has a right to defend its patent against cancellation by the USPTO even after expiration — and up to the six-year timeline for back-damages.

Setting procedure aside. On the merits, the majority agreed with the patentee Sony – that the PTAB had erred in claim construction.  Here, this is a case where means-plus-function language helped the patentee with a narrow claim construction.

The patent here appears to be fairly broadly drafted — claiming an “information reproducing device” that is able to play “a plurality of voice data, each voice
data having similar contents translated into different languages are multiplexedly recorded as audio data of plural channels.”  The device has a “default code” for which language to play.

The argument on appeal involved the term “reproducing means for reproducing the audio data of the channel designated by the default value stored in the storing means.”  As a means-plus-function term, the element was interpreted on appeal under 112(f):

(f) Element in Claim for a Combination.— An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.

Here, although you might imagine that the “reproduction means” is some piece of hardware, the patent only describes it in software form.  The result is that the best interpretation of the claim term also limits it to software form — since means plus function limitations are construed only “to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.”  The prior art described the claimed means in hardware form rather than software. On remand, the PTAB will need to determine whether the claims remain obvious in light of that narrowed claim construction.

= = = = =

Important quote from the case:

We are not bound by the parties’ arguments as to claim construction. See Exxon Chem. Patents, Inc. v. Lubrizol Corp., 64 F.3d 1553, 1555–58 (Fed. Cir. 1995) (“[T]he [court] has an independent obligation to determine the meaning of the claims, notwithstanding the views asserted by the adversary parties.”).

Means-Plus-Function Claims Lose at the Federal Circuit

Articles and Publications, Patent

35 U.S.C. Section 112 paragraph 6 provides a statutory basis for drafting patent claims in what is known as means-plus-function language.  The language allows a patentee to write a claim based on the function of a device rather than its actual structure.  This language is, in the words of Gene Quinn, "dangerously seductive."  Means claims are easy to write, and there is a belief that the means language adds some "fuzzy math" to the court’s calculation of infringement.

Over the past several years, the Federal Circuit has continually limited the scope of means limitations.  Such a limitation is construed by the court to cover the corresponding structure as described in the specification and equivalents thereof.  Thus, if you use means claim limitations, you should make absolutely sure that the structure is completely (and broadly?) described in the specification.

In an admittedly unscientific survey of case law, attorney Hal Wegner searched through the last fifteen CAFC opinions that dealt with the interpretation of means-plus-function language. Results:

Patentee’s construction of "means" term accepted: 0%
Patentee’s argument that term was not a "means" term accepted: 14
Summary judgment of non-infringement reversed: 7%
Losses for Patentee: 80%.

Thus, the means language was a loser for the patent holder in 80% of cases. (12 of 15 cases). (Hal, thanks for conducting the study).

Bottom Line: MPF claims are risky.  They are fine to add as insurance after you have drafted more descriptive claims, but they should not be trusted. 

If you are still writing MPF claims, you are not alone — thousands of patents using such language are issued annually.