By Dennis Crouch
Apple Inc. v. Samsung Electronics Co., Ltd., — F.3d —, 2012 WL 4820601, App. No. 2012-1507 (Fed. Cir. 2012) (N.D.Ill. 12-CV-0630) (APPLE II)
This is my second post on the appellate decision that I term Apple II. [Read the First Apple II Post.] In Apple II, the Court of Appeals for the Federal Circuit vacated N.D. California District Court Judge Koh’s preliminary injunction order that would have forced Samsung’s Galaxy Nexus smartphone off the market pending conclusion of the infringement lawsuit. The appellate court provided two independent reasons for vacating the preliminary injunction order: (1) that Apple had failed to provide proof of irreparable harm due to Samsung’s continued use of the particular invention covered by the patent; and (2) that a finding of infringement is unlikely based upon the claim construction as corrected by the Federal Circuit. This post focuses on the claim construction issue.
The patent at issue in this case is U.S. Patent No. 8,086,604. That patent issued in December 2011 but stems from upon an application originally filed in January 2000 and appears to be largely directed to aspects of Apple’s Siri application (although the original invention is not tied to a mobile device). After construing claim 6 of the patent, the district court found that Samsung’s product likely infringes. On appeal, the Federal Circuit has reversed that claim construction in a way that substantially narrows claim scope and thus negates any likelihood that Samsung infringes. Although the claim-at-issue uses open-ended “comprising” language, the appellate panel held that an apparatus that includes additional similar (but non-infringing) elements would not be covered by the patent.
Claim 6 is directed to an “apparatus for locating information in a network” comprising various software elements. The basic idea is that the system has a set of heuristic modules corresponding to different search areas. We all use heuristics for searching. For example, if you want to find prior art relating to a chemical compound, you might hire an expert chemistry searcher. On the other hand, if your search area is airplane seat design, you might hire someone else. Under the language of the patent, these search strategies could be defined as two different heuristic modules each corresponding to their respective search area. In the lawsuit itself, Apple identified the three heuristic modules as (1) a Google Search Module; (2) a Module that allows for Search of Internet Browsing History; and (3) a Search of a user’s Contacts List. One difficulty with considering this case is that this invention appears in hindsight to be ridiculously obvious. However, obviousness is not the topic of the appeal.
The patent claim requires “a plurality of heuristic modules … wherein: each heuristic module corresponds to a respective area of search and employs a different, predetermined heuristic algorithm corresponding to said respective area.” Samsung’s product allegedly uses a plurality of heuristic modules corresponding to different search areas. However, the accused product does not use “different, predetermined heuristic algorithm” for each and every different search area. I have modeled this below with a diagram of the Galaxy Nexus Search System associated with five different Search Areas (A,B,C,D,E) but only three different Heuristic Modules (X,Y,Z). Although some of the Heuristic Modules are re-used by the system, Apple argues that the system still infringes under standard patent law that would find the addition of non-infringing elements to an infringing core does not negate infringement. See CIAS, Inc. v. Alliance Gaming Corp., 504 F.3d 1356, 1360 (Fed. Cir. 2007) (infringing device may include additional, unrecited elements).
Writing for the unanimous panel, Judge Prost (joined by Judge Moore and Reyna) rejected Apple’s argument and instead agreed with Samsung that the claim “requires that [each and] every heuristic module within the accused device use a unique heuristic algorithm.”
[D]espite the use of “comprising,” claim 6 is not amenable to the addition of other modules that do not use different heuristic algorithms because such addition would impermissibly wipe out the express limitation that requires every module to have a unique heuristic algorithm.
It appears from the court’s remarks that a simple claim amendment might have saved Apple’s case:
- Original: “a plurality of heuristic modules … wherein: each heuristic module corresponds to a respective area of search and employs a different, predetermined heuristic algorithm …”
- Proposed: “a plurality of heuristic modules … wherein: each of the plurality of heuristic module corresponds to a respective area of search and employs a different, predetermined heuristic algorithm …”
However, Apple might not have been able to obtain patent protection if it had clarified that the system might use the same heuristic algorithm for a variety of search areas. In particular, during prosecution Apple used the limitation in question to avoid prior art. The company argued “Andreoli does not describe … that each of the local and remote search operations employs a different heuristic algorithm.” Although the appellate panel based its decision on what I see as odd linguistics, this prosecution history argument is probably the most damning evidence.
Apple has at least one continuation application pending and thus has the opportunity to further shape its claims in a way that captures the market while avoiding prior art. See U.S. Patent Application Pub. No. 2012-0166477. The USPTO is also in the early stages of conducting an inter partes reexamination that was filed in August 2012 by Motorola Mobility (Google). See Inter Partes Reexamination No. 95/000,685.
I was talking about the previous years of prosecution and litigation leading up to this “easy first office action.”
How does the proposed wording remedies the flows of the original? seems to me they both convey the same…
The attorney merely called them what the Examiner called them. I guess the attorney could have said “they” use the same heuristics but chose to say the local and remote use the same.
But the local and remote locations are not the “same location” (as you seemed to indicate in your 7:07 comment).
the reality is that I have a large portfolio that I feel has been weakened by the CAFC for no good reason.
I think this case can be limited to its facts for the very reasons we are discussing. If your portfolio was drafted in the same form as these claims then it was already “weakened” for the same reason that these claims were “weakened”, i.e., they could be more clear or at least an unambiguous clarification could be made in the prosecution history.
It’s like those [oldsetp]+[newthought] claims that got chopped down in Prometheus. They were *always* problematic. All that was needed was for someone to call the patentee’s bluff. That said, rest assured that a lot of money was made off that g-rb-age and probably continues to be made because people are just too lazy to take a stand or too clueless to recognize the issue even after the court case.
The attorney merely called them what the Examiner called them. I guess the attorney could have said “they” use the same heuristics but chose to say the local and remote use the same. Since that was all the Examiner cited (two) it was also true to say that “each” heuristic is the same. That should not mean that the history limits the claim as the CAFC said. Look at the rejection and then honestly tell me you would have responded differently (without this decision)
I still say the claim as written unambiguously covers the examples you cite. Obviously the panel disagrees with me. You could say at least two (pluality) modules where each of the at least two modules correspond to different areas and each of the at least two modules employs different heurisitcs. But why would that change the outcome?
I am not on the court but I manage thousands of patents and the drafter did a good job. I would have prefered that the claim be drafted as Dennis noted above but I don’t see that form trumping substance. What was claimed is the inner circle of the diagram. I have always been told, and therefore taught, that adding additional elements outside of that circle does not change infringement.
This is fun to debate but the reality is that I have a large portfolio that I feel has been weakened by the CAFC for no good reason. They could have simply said that the prosecution history raises enough doubt to deny injunction. This would have allowed the district court to investigate the history and determined what the applicant and art really said.
“The patent at issue in this case … stems from upon an application originally filed in January 2000 and appears to be largely directed to aspects of Apple’s Siri application.”
That would have to be an accident/coincidence, since Apple did not acquire Siri until 2010. Siri was originally a start-up founded by SRI International, that received/licensed IP from SRI and further developed its own technology. Apple no doubt did much to further improve it, but did not touch Siri until a decade after filing the above patent application.
This article is very much informative. These days, people should be a lot more concious and aware of whats going on with the leaders in communication such as Apple and Samsung (as mentioned in this article). Thanks for sharing this post.
Sounds like it will be an easy first office action rejection — and not an INCREDIBLE waste
what is the effective date of the application?
It looks like the Examiner said that the art teaches two heuristics that search the same location
That’s interesting. Makes one wonder why Apple would respond to Andreoli by referring specifically to “local and remote search operations” when those terms are not recited in Apple’s claim. Very curious.
I always find it odd that some choose to quibble over matters such as these. Yes, you don’t believe the statement is necessarily contradictory with the narrow construction proposed during litigation. That’s great! But Dennis found it noteworthy. And it seems contradictory to me. And look what happened at the Federal Circuit. The bottom line is that patent prosecutors are paid a lot of money for their writing skill and when they fail to convey in unambiguous terms the breadth of their invention and exactly how it differs from the prior art, there is a probably an easy-to-understand reason for that failure.
“wherein at least two of said plurality of modules employ distinct heuristic algorithms.” Makes no sense.
It makes plenty of sense.
Why would you say at least two when a pluraity by definition is at least two?
Because my claim unambiguously covers the following situations:
only two modules, both modules with different algorithms
ten million modules, with each module using a unique algorithm
ten million modules, where all the modules except one use the same algorithm
That’s why. An alternative you might prefer : “wherein said plurality of modules do not employ identical heuristic algorithms,”; or (relatively clunky but still crystal clear) “wherein said plurality of modules comprises a first and a second module, wherein said first module employs a heuristic algorithm which differs from a heuristic algorithm employed by said second module.”
I would not have anticipated the CAFC turning basic claim construction on its head.
Are you kidding me? For cripes sake, there is at least one judge on the CAFC who appears to believe that broad claim terms in independent claims should be limited to read only on embodiments disclosed in the specification, even when those narrower embodiments are recited in dependent claims! Never underestimate what your opponent in litigation will try to persuade a judge is “correct.”
MM, I looked at the reference and PTO rejection. It looks like the Examiner said that the art teaches two heuristics that search the same location where one heuristic searches the location differently from the other heuristic. I dont see how this limits the claim. The claim has two modules corresponding to differnt locations and employing two heurisitcs.
If I can find Andreoli and understand it (not my area) maybe I will follow-up. My position is that if Andreoli teaches 100 different modules but they all use the same heuristics then there is no limiting statement because any plurality (at least two) of modules all use the same algorithm. So a claim that says at least two module wherein each of the module use different search algorithms is not in conflict.
Again, the statement in prosecution may not be that damaging unless ther are two modules using two different heurisitcs. I would not punish my attorneys if they said the same thing. I would not have anticipated the CAFC turning basic claim construction on its head. Maybe I should.
“wherein at least two of said plurality of modules employ distinct heuristic algorithms.” Makes no sense. Why would you say at least two when a pluraity by definition is at least two?
a plurality of heuristic modules (A&B) … wherein: each heuristic module (A&B) corresponds to a respective area of search (A&B) and employs a different, predetermined heuristic algorithm (A&B).
So, module A coresponds to a search A and uses heuristic A. Module B coresponds to search B and uses heuristic B.
“Andreoli does not describe … that each of the local and remote search operations employs a different heuristic algorithm.”
A rebuttal that is consistent with the patentee’s preferred construction would read: “Andreoli does not describe that different heuristic algorithms are used by at least two of the recited modules.”
I don’t buy the prosecution history argument without more information.
Let us know what you find. I’m pretty sure you’ll be disappointed. Not that it matters. This patent was pretty much pffffffffft on arrival. Heckuva job, PTO.
The claim is simply to at least two modules each having different heuristic.
Then the claim should have been written: “wherein at least two of said plurality of modules employ distinct heuristic algorithms.” Very straightforward.
Instead, the claim was written in a clunky, confusing fashion, almost as if the applicant was trying to get a claim that could be interpreted in two very different ways, at the whim of the patentee and almost surely depending on the nature of the alleged infringer’s product/actions.
Then of course there was the statement made during prosecution, noted by Dennis above, which directly contradicts the claim construction proposed by the patentee during litigation but which is completely consistent with the Federal Circuit’s interpretation.
I don’t buy the prosecution history argument without more information. Did the cited art have multiple modules but they all used the same heuristics (like the inside circle of diagram but the same search)? If so then there is no contradiction with the claim. If there were at least two modules in the prior art using two different heuristics (as in the diagram) then the claim is invalid.
Ned, I disagree. The term “each” clearly refers to the claimed plurality. The claim is simply to at least two modules each having different heuristic. Why is that interpretation wrong?
“…each heuristic module corresponds to a respective area of search and employs a different,
predetermined heuristic …to a respective area of search…”
Areas: internet, address book, browsing history.
The Fed’s opinion is on point.
Apple’s construction would reduce the claim to a claim simply requiring more than one heuristic.
One begins to suspect that a certain lower court Judge is not all that sharp.
This CAFC decision is very similar to a decision by Judge Newman reversing a BPAI decision using the same logic. The BPAI said that the prior art only had to show that each of a set of X met limitation Y. The CAFC said that each element for the entire product (a shoe?) had to meet limitation Y.
I think both this decision and the Newman decision were decided wrongly. They don’t seem to understand how the term “comprising” works with set theory / Venn diagrams.
Come to think of it, it sounds a lot like what Windows would do if your system needed a printer or video or mouse driver. First it would look on your hard drive, then it would look on removable media, like CDROM or Floppy disk and failing those, it would ask if you wanted it to look on the interweb….
It sounds a lot like what “google desktop” was doing some years ago. But I don’t think that predates 2000.
In particular, during prosecution Apple used the limitation in question to avoid prior art. The company argued “Andreoli does not describe … that each of the local and remote search operations employs a different heuristic algorithm.” Although the appellate panel based its decision on what I see as odd linguistics, this prosecution history argument is probably the most damning evidence.
Nobody could have predicted that the applicant would make such an argument during prosecution and then say the opposite during litigation! A classic and classy maneuver.
What’s really weird is that such an argument could ever suffice to move prosecution forward. Who cares if one reference doesn’t teach the use of different heuristic algorithms for searching different databases? Different heuristic algorithms existed at the time of filing. Everybody knew that. The skilled artisan would be motivated to use different heuristic algorithms for searching different databases because certain heuristic algorithms were known to provide better results than others depending on the nature of the database being searched. Done.
That prosecution history is truly silly but it still doesn’t top the prosecution argument that keeping a log of what ads are skipped is COMPLETELY DIFFERENT from keeping a log of what ads are watched. That one is still the all-time champion (although I’m sure similarly inane arguments are made on a daily basis).
One difficulty with considering this case is that this invention appears in hindsight to be ridiculously obvious
Truly. But you can already hear Apple’s attorneys barking about those “secondary factors”,. e.g., all those sales and the “long felt need” (“if it was so obvious why wasn’t it invented earlier, huh?huh?huh?”).
In any case, in the continuation the PTO is citing 6,070,158 (“Kirsh”) as an anticipatory reference under 102(e). Even if some “technical” argument is made to get around Kirsh, the obviousness argument should remain insurmountable.
What an incredible waste of judicial and USPTO resources.
I was just reading the continuation application that was filed in November 2011. Claim 1 is broadly directed to a search algorithm that searches a both local files and internet accessible files. Its hard for me to believe that the attorney who filed this had a good faith belief that the claims are patentable.
Claims 1: A computerized method, comprising
receiving at least a partial search query;
storing the at least partial search query on a non-transitory computer readable medium;
searching a first plurality of files for matches with the at least partial search query, wherein the first plurality of files is located across an internet connection coupled to the non-transitory computer readable medium;
searching a second plurality of files for matches with the at least partial search query, wherein the second plurality of files is only available via local access from a computer that includes the non-transitory computer readable medium; and
receiving a notification of at least one search result.
Do you have prior art? link to patents.stackexchange.com
Comments are closed.