« Peers on Hiatus | Main | Lever »

July 20, 2009

Off the Shelf

Saint-Gobain sought declaratory judgment of noninfringement and invalidity over refrigerator shelf patents owned by competitor Gemtron. Gemtron counterclaimed and ended up as plaintiff. The dispute came down to a single limitation in a single claim, over when the claimed invention had the glass panel shelf bottom snap into the frame. The answer: no set time.

Gemtron v. Saint-Gobain (CAFC 2009-1001) precedential

Claim 23 of 6,679,573, a continuation of 6,422,673, was the disputed claim.

Unlike prior art shelves that were constructed by bonding the edge of the glass panel to the frame using adhesive, the claimed shelf secures the glass panel in the frame using "relatively resilient" fingers so that the glass panel is "snap-secured" into place. Id. col.2 ll.42-47.

Saint-Gobain does not dispute that its accused refrigerator shelves meet all of the limitations of claim 23 except the limitation requiring that the frame have the claimed "relatively resilient end edge portion." S.J. Op. at 6. The district court construed the claim term "relatively resilient end edge portion which temporarily deflects and subsequently rebounds to snap-secure" to mean that "the end edge portion is sufficiently resilient that it can temporarily deflect and subsequently rebound when glass is being inserted into the frame." Id. at 2-3.

The only disputed claim term is "relatively resilient end edge portion which temporarily deflects and subsequently rebounds to snap-secure." '573 patent col.8 ll.61-62. The district court construed this term to mean that "the end edge portion is sufficiently resilient that it can temporarily deflect and subsequently rebound when glass is being inserted into the frame." S.J. Op. at 3 (emphasis added). Saint-Gobain argues that the term should have been construed to mean that "the end edge portion is sufficiently flexible to permit the glass in the finished product to be pushed out of the frame and pushed back into the frame." Br. for Defendant-Appellant Saint-Gobain Corp. at 25 (emphasis added). More particularly, Saint-Gobain argues that "relatively resilient" should not mean "temporarily resilient immediately after cooking in an oven and before any opportunity to cool." Id. at 26. Gemtron argues that the district court's construction was correct. The parties do not dispute that the "relatively resilient" limitation requires that the frame be flexible at some point in time. The dispute is about precisely when the frame must be flexible to satisfy the "relatively resilient" limitation.

Because neither party has identified any portion of the prosecution history or any extrinsic evidence bearing on the claim construction issues before us, we confine our analysis to the claim language and the specification. Looking first to the claim language, we note that claim 23 has no express temporal or temperature limitation requiring that the end edge portion be relatively resilient "always," or "at all temperatures," or "when in use in a refrigerator."

Like the claims, the specification focuses on the characteristics of the frame that enable snap-secure assembly of the claimed shelf... In fact, every time the structure of the "relatively resilient" edge portions is mentioned in the specification, it is in the context of a discussion of how that structure functions while the shelf is assembled.... There is no discussion in the specification of any purpose for or value of the "relatively resilient" structural characteristic of the end edge portions of the frame, other than to facilitate assembly of the shelf. This indicates that the end edge portions of the frame have the claimed structural characteristic--"relatively resilient"--if they are able to deflect at the time the shelf is assembled, to "snap-secure" the glass panel within the frame.

For these reasons, we affirm the district court's construction of "relatively resilient end edge portion which temporarily deflects and subsequently rebounds to snap-secure" as used in claim 23 to mean that the end edge portion must be sufficiently resilient that it can temporarily deflect and subsequently rebound when glass is being inserted into the frame.

Desperate thrashers in patent litigation always trot out some mangy dog of an argument in fond hopes of bamboozling the court. For Saint-Gobain, the ploy was to try to package the disputed limitation as product-by-process. The CAFC didn't bite.

Saint-Gobain argues that a construction requiring only that the frame deflect at the time that glass is being inserted into the frame during assembly transforms the "relatively resilient" limitation of claim 23 into a product-by-process limitation. We disagree. The limitation requires that the glass panel be "snap-secure[d]" in the frame. "Snap-secure[d]" describes the structural relationship between the glass panel and the frame, which is possessed by the claimed shelf because of the structural characteristics of the individual components. See, e.g., 3M Innovative Props. Co. v. Avery Dennison Corp., 350 F.3d 1365, 1371 (Fed. Cir. 2003) ("[E]ven words of limitation that can connote with equal force a structural characteristic of the product or a process of manufacture are commonly and by default interpreted in their structural sense . . . ."); Hazani v. U.S. Int'l Trade Comm'n, 126 F.3d 1473, 1479 (Fed. Cir. 1997) (holding that claim is not product-by-process claim if it "describes the product more by its structure than by the process used to obtain it"); see also Garnero, 412 F.2d at 279 (noting that past-tense verbs such as "'intermixed,' 'ground in place,' 'press fitted,' 'etched,' and 'welded,' all . . . at one time or another have been separately held capable of construction as structural, rather than process, limitations."); Eric P. Mirabel, Product-By-Process Claims: A Practical Perspective, 68 J. Pat. & Trademark Off. Soc'y 3, 4-7 (1986). Defining a structural component by its functional as well as its physical characteristics is different from defining a structure solely by the process by which it is made. Our construction therefore does not transform the "relatively resilient" limitation of claim 23 into a product-by-process limitation.

The last gasp for Saint-Gobain was to wriggle for noninfringement because the accused product was made in Mexico. Never mind the law.

Claim 23 is directed to an apparatus--"[a] refrigerator shelf"--not a process. '573 patent col.8 l.32. Under 35 U.S.C. ยง 271(a), an accused infringer infringes an apparatus claim if it "makes, uses, offers to sell, or sells" the claimed apparatus "within the United States," or "imports [the apparatus] into the United States." Critically, it is the infringing act--making, using, offering to sell, selling, or importing--that must be within (or into) the United States. Even if an infringing product is manufactured outside of the United States, a person infringes if he imports the product, or uses, offers to sell, or sells it in the United States. See, e.g., In re N. Pigment Co., 71 F.2d 447, 456 (CCPA 1934) ("It has long been settled that articles patented in the United States cannot be manufactured abroad, imported, and sold in violation of the rights of the patentee.").

Affirmed. Patent valid and infringed.

Posted by Patent Hawk at July 20, 2009 9:49 PM | Claim Construction

Comments

"relatively resilient"

Saint-Gobain had a large pair in bringing up 35 USC 271.

Emphasis on had.

It's interesting that the court seemed to ignore the adverb "relatively" which described resilient. Relatively means "sufficiently, not absolute" or "enough to be suitable." I don't think the edge portions did not lose their relatively resilient characteristic just because they were cooled.

Posted by: niRPa at July 21, 2009 3:06 AM

Arggh. Sorry. I don't think the edge portions lost their relatively resilient characteristic just because they were cooled.

Posted by: coffeeless niRPa at July 21, 2009 3:08 AM

"Relatively means "sufficiently, not absolute" or "enough to be suitable."

Imma readin ma dictionary and I don't see that definition in this english language.

What the word means is simple.

http://www.thefreedictionary.com/relatively

And it means precisely that in all circumstances.

It's every dictionary in the world against you Nirpa.

http://www.merriam-webster.com/dictionary/relatively

Such terms, when used and don't have anything specified, leave the claim open to 112. Every. Single. Time.

If the court was relying on the presumption o validity to come to this decision they should have mentioned it.

Posted by: 6000 at July 21, 2009 12:01 PM

btw nirpa, their argument isn't that their edge wasn't resilient, or ""sufficiently, not absolute" or "enough to be suitable"" resilient. The argument they made was that when they cooled the edges were no longer capable of performing the method step/functional limitation that comes after the resilient limitation. And they are right. Problem is, if the product meets the limitation at any time then it would "infringe". However, like they said, it only infringed when it was in Mexico, and by the time it got here the structure no longer met even the bogus limitations dreamt up by the DC much less meeting the actual limiations of the claim.

They were correct on all accounts but failed to properly frame the argument they were making.

Posted by: 6000 at July 21, 2009 12:07 PM

"It's every dictionary in the world against you Nirpa."

You mean every free on-line dictionary, I trow. And??

I thought they reached the proper result in the case. The claim clearly covered the snap-secured connection. The connection was precisely described in the claim by the manner in which the connection can be distinguished from other connections (i.e., there is a "snap" during assembly) - the court said it wasn't a product-by-process claim (i.e., they were not claiming the actual snapping, but rather the connection; nonetheless, the result would have been the same even if it were construed as a product-by-process).

The PTO uses the same snap jargon to distinguish products i.e., connections, not methods of assembly, in Class 24, Subclasses 66.1 and 270.

Clearly Saint-Gobain had the snap-secured connection.

Posted by: niRPa at July 21, 2009 3:41 PM

"The claim clearly covered the snap-secured connection."

Perhaps, but even if we give you that, it had to be that same "snap-secured" connection when it was in the US. It wasn't.

"The connection was precisely described in the claim by the manner in which the connection can be distinguished from other connections "

There are much easier, and less indefinite ways of distinguishing it. There is also a way to make their method step limitation into a functional limiation. They choose not to use any of those ways.

" the court said it wasn't a product-by-process claim "

No, it isn't a product BY process, it's a product with manner of using the product claim. Even worse than a product by process. Completely non-statutory.

"if it were construed as a product-by-process'

That would be blatantly improper, but if it was construed thus, doesn't the recent case on pbp state that the method limitations would have to be performed in order to infringe? Those method limitations were CERTAINLY not (and not even allegedly) performed in the US. The only way to finangle this decision into infringment was just in the manner they did.

"The PTO uses the same snap jargon to distinguish products i.e., connections, not methods of assembly, in Class 24, Subclasses 66.1 and 270."

Ok, perhaps there is some art mandated standard at work here, but if so there should have been some evidence of such on display rather than all this nonsense that got spouted.

I don't know, but I'm let down that mah boys would mess this one up. But, hey, everyone has a bad day.

Posted by: 6000 at July 22, 2009 12:51 PM

"Perhaps, but even if we give you that, it had to be that same "snap-secured" connection when it was in the US. It wasn't."

Was too.

"There are much easier, and less indefinite ways of distinguishing it."

Hindsight is 20-20.

"No, it isn't a product BY process, it's a product with manner of using the product claim. Even worse than a product by process. Completely non-statutory."

Well product-by-process claims and obviousness-type double patenting rejections are completely non-statutory too, so that really doesn't phase me. There is still the presumption of validity and the legal principles of claim interpretation that you must live with: regardless of all the "what it does" and "what it was" clauses in a product claim issued by an Examiner, the court must construe it as a product: "what it is." That is proper claim interpretation.

"That would be blatantly improper, but if it was construed thus, doesn't the recent case on pbp state that the method limitations would have to be performed in order to infringe?"

The method was performed by Saint-Gobain in Mexico. I assume the court would have held (had they so construed the claim as a product-by-process) that the "process" part of the claim is afforded the benefit of 271(g), and hence performance in Mexico of the "process" part constitutes infringement in the U.S. when the product is imported.

When looking at this case it's helpful to see what problems would have arisen if the court went the other way:

For example, consider a patented electrical connector claiming resilient contacts. The contacts need to be resilient to make the initial electrical connection and provide bias for maintaining the electrical connection. The same connector with resilient contacts could be assembled in a circuit board made in Singapore, a drop of adhesive (or a removable immobilizing shim) applied to the contacts to make them relatively rigid, and imported into the U.S. in that condition without infringement. Or so would say Saint-Gobain.

"But your honor, our connectors are no longer resilient - we glued them to make them rigid after we assembled them!"

No. They still provide bias - they are still relatively resilient. (And they functioned as claimed when you assembled them too.)

Posted by: niRPa at July 22, 2009 3:17 PM

"Was too."

You didn't read the case very well, or perhaps it was the claim you didn't read very well. The product, at the time it entered the good ol' us of a, was not capable of performing the intended use/function/method step language.

"Hindsight is 20-20."

I agree, it is the attorney's/patentee's job to see 20-20 from the get go. And in this case, it wasn't very hard to see the proper claim format from the get go, one must therefor assume that the patentee deliberately used the language that he did in order to be indefinite and perhaps capture more scope.

"Well product-by-process claims and obviousness-type double patenting rejections are completely non-statutory too, so that really doesn't phase me."

Let me restate my objection to the present claims, their type is completely non-statutory and there has been no judicial exception carved out for them as of yet.

"There is still the presumption of validity and the legal principles of claim interpretation that you must live with: regardless of all the "what it does" and "what it was" clauses in a product claim issued by an Examiner, the court must construe it as a product: "what it is.""

If they were relying on the presumption, then they should have at least paid it some lip service. That's all I'm saying.

"The same connector with resilient contacts could be assembled in a circuit board made in Singapore, a drop of adhesive (or a removable immobilizing shim) applied to the contacts to make them relatively rigid, and imported into the U.S. in that condition without infringement. Or so would say Saint-Gobain."

From what I can tell from your crafted hypo, the shim was inserted before they arrived in the US (and presumably the shim is left in). Therefor, when they come across the border they are resilient (because the shim is in) and infringe. That is the complete arse backwards situation from the one in the actual case. An analogous situation to the one in the actual case would involve the shim being put in, and then taken out, and then the product coming across the border sans shim. When they came across the border, they do not infringe.

That's one of the reasons I don't think you read the case very well. When the product came across the border, it was no longer relatively resilient enough to snap into place or whatever method step limitation they chucked in there.

"they are still relatively resilient."

They aren't compared to my rock hard arms, and they probably never were. Therefor, they probably never infringed, even in Mexico. I think you get my point. We don't know how resilient they had to be to infringe. POSHITA doesn't, and no member of the public does. This is the part when 112 comes to town and starts a ruckus.

In any event, I think you see where I'm coming from and I see where you're coming from. So like you said, then it is helpful to see how the court would have gone had it have gone the other way.

First, Linn would have asked the guy what the resilience which is required to infringe is relative to, and the patentee wouldn't have an answer. Then the Linn would say ok, 112 might be a problem, I can't make heads or tails of this limitation, the patentee can't, none of the lawyers can, and neither could one of ordinary, extraordinary or less than ordinary skill. And then him n the boys would decide if a claim that doesn't meet 112 is invalid. They do this all the time.

But of course, validity is a question reserved for a court, and a question which I have no feelings on personally. Claims may very well not have to meet the conditions of 112 in order to be valid.

At the end of whatever decision they wrought, they'd throw in a "is that a method step limitation chillin out in your product claim? LOL" Then they might impose some rule 11 sanctions on the plantiff's lawyers. And everyone would go home happy :)

In a perfect world, I'm just saying, ok?

Posted by: 6 at July 22, 2009 6:45 PM

"The claim clearly covered the snap-secured connection."

No. It covers "a relatively resilient end edge portion which temporarily deflects and subsequently rebounds" -- not "a relatively resilient end edge portion which temporarily deflected and subsequently rebounded."

At the time of importatation and ever since, it never "deflects and subsequently rebounds" ... no infringement here.

If the claim had said, snap-secured ... okay, but it said the edge portion actually deflects and rebounds. It didn't at the time of importation. It didn't afterwards.

Posted by: anonymous at July 23, 2009 10:32 AM

"importatation" is a new word

if the claim said the edge portion "connects" the frame to the shelf using a snap fit, then you could argue that the verb "connects" reads on the device because the connecting was occuring in the device that was imported

there was no deflecting or rebounding that was occuring in the imported device

Posted by: anonymous at July 23, 2009 10:37 AM

Person of ordinary skill and having common sense would read the claim as "snap secured" and thus it is infringed.

Posted by: John Prosecutor at July 23, 2009 2:35 PM

The claim language:

"a relatively resilient end edge portion which temporarily deflects and subsequently rebounds to snap-secure one of said glass piece front and rear edges in the glass piece edge-receiving channel of said at least one front and rear frame portion."

MPEP 2114:

"[A]pparatus claims cover what a device is, not what a device does." Hewlett-Packard Co. v. Bausch & Lomb Inc., 909 F.2d 1464, 1469, 15 USPQ2d 1525, 1528 (Fed. Cir. 1990)

Regardless of how you may want to construe the claims (e.g. you say they are claiming what the device does - i.e. the deflecting and rebounding), you must construe them properly (i.e. for what the device is). That is part and parcel of the presumption of validity - you may not construe the claim to be invalid when a construction that is valid is (equally) plausible.

"there was no deflecting or rebounding that was occurring in the imported device"

But there was the snap-secured connection - you cannot deny that (o.k. you can). And there was an end edge portion that was not absolutely and completely rigid and hence it was relatively resilient and could both deflect (e.g. by 100 microns) and rebound.

P.S. I bet 6K can't find that importatation word in his dictionary. But if you used it in a claim (instead of importation), we would all construe it for what you meant so as to render the claim valid. (Spelling errors do not affect claim validity unless they make the claim indefinite.)

6K: regarding the word resilient. You are saying your "rock hard" arms are relatively resilient? I have no problem believing that. So is sponge rubber, a sheet of paper, a paperclip, the screen on my LCD monitor. Applesauce is not relatively resilient.

Posted by: niRPa at July 23, 2009 4:00 PM

P.S. 6K I'm right there with you on the "broadest reasonable interpretation" stuff. I have to think that any attorney who has ever been an Examiner for a significant period must understand why it's necessary. I remember one (very good) attorney complaining (very clearly) about it when I was an examiner... I wish I could tell the story.... but BRI carried the day.

I do believe it is necessary for proper examination of arguably broad claims. If you could find one ex-Examiner that honestly believes otherwise I'd be surprised.

Posted by: niRPa at July 23, 2009 4:14 PM

"Regardless of how you may want to construe the claims (e.g. you say they are claiming what the device does - i.e. the deflecting and rebounding), you must construe them properly (i.e. for what the device is)."

No, the claim dies under 101 at the outset. It isn't an apparatus/product claim, because it covers what the device does. It isn't a composition, and it for sure ain't a method. No stat category? Goodbye claim.

There is no "let's ignore the limitation in an issued claim because it might kill the claim, and then make up some different limitations to assist the claim in differentiating itself from the prior art" doctrine that I'm aware of.

"you may not construe the claim to be invalid when a construction that is valid is (equally) plausible."

I agree, too bad the other "construction" is a figment of your imagination and isn't "plausible" at all.

What's next? Shall we start to construe the claim:

1. A bicycle.

To mean 1. A poly-carbon-nanotube chain with the following structure: XYZ?

The later would plausibly be a valid claim. But we don't go making up claims.

This claim doesn't cover anything magical, and there is no reason to invoke magic to make up an acceptable claim.

And even if, if, you wanted to do such sorcery in the courtroom, you need to invoke your magic spell by saying the incantation: Well, according to the presumption of validity we can't read the claim as written, we need to make something up so that it is valid.

And that's all I'm saying about that. They didn't invoke their magic spell. Courts can go results oriented any time they please, they do it every day, but they need to invoke some magic to make it proper.

"the snap-secured connection"

No, there was the action of snapping secured. That isn't the snap-secured connection.

Plus, and I don't think you picked this up as well, in the "infringing" device, there was no snapped secure connection because the refrigerator hadn't been put together yet. We're talking about the piece with edges by itself, unconnected to anything.

The case is this iirc. The "infringers" made a product that happened to be able to perform some claim language while it was in mexico and the product was under construction, but they never actually used it in such a manner, either in mexico, or the US. That is, they made this insertable piece, but they didn't install it in a fridge.

It's straight up extortion/cockblockery is what it is, and I'm let down that my boys would allow it. But, I make mistakes too.

"I do believe it is necessary for proper examination of arguably broad claims. If you could find one ex-Examiner that honestly believes otherwise I'd be surprised."

Indeed, the attorneys that don't like it don't really understand that it is, at heart, a necessary evil. If it is an evil at all.

"And there was an end edge portion that was not absolutely and completely rigid and hence it was relatively resilient and could both deflect (e.g. by 100 microns) and rebound."

The last thing I don't think you quite get, even if the court did adopt your magical explanation (theirs was lots different than yours), and your definition of resilient, and deflection, then your claim probably doesn't even define itself over the prior art anymore. You gotta remember, you're fin around with their point of novelty.

"You are saying your "rock hard" arms are relatively resilient?"

No, they're absolutely and completely resilient. Nothing made of plastic (the accused infringing device) is (relatively) resilient relative to them because they are the absolute that can ever be achieved.

My point is this, if you want to say "substantially" or whatever, then fine, say that. If you say relatively, then you mean relatively and just the same as I crucify other apps for this same mistake, there shouldn't be an exception made for claims that do the same thing just because they slipped past some examiner who wasn't paying attention.

Bottom line, I know you want to bend over backwards to save a "valuable" claim. But that isn't what the presumption is about.

Posted by: 6 at July 24, 2009 2:45 PM

Hey, 6 et al.,

If the claim were amended as follows:

"a relatively resilient end edge portion which SUBSEQUENT TO ASSEMBLY temporarily deflects and subsequently rebounds to snap-secure one of said glass piece front and rear edges in the glass piece edge-receiving channel of said at least one front and rear frame portion,"

would you view that as a narrowing amendment?

How about:

"a relatively resilient end edge portion which AT THE TIME OF ASSEMBLY temporarily deflects and subsequently rebounds to snap-secure one of said glass piece front and rear edges in the glass piece edge-receiving channel of said at least one front and rear frame portion?"

What I'm getting at is that the claim language does not expressly state the time at which the deflection and rebounding occurred. If those clasues had been dependent claims, would you then agree that the independent claim, based on claim differentiation, must be broad enough to cover both cases? If so, then why do you insist that, absent the existence of those dependent claims, the claim must be interpreted so narrowly as to exclude one of those cases?

Surely, if you were examining such a claim and found some art that showed the deflecting and rebounding at time of assembly, you would have required the Applicant to amend the claim to specify a limiation to being able to deflect and rebound post assembly. So I ask you, why would you be willing to read the limitation so broadly during prosecution, but not at time of enforcement?

I think you have an overly adversarial mindset.


Posted by: anon at July 27, 2009 10:27 AM

"would you view that as a narrowing amendment?"

Not really, how would an edge temporarily deflect an subsequently rebound prior to, or during assembly of the edge? The "edge" as such doesn't even exist during its own assembly. What is "assembly" refering to? Assembly of the edge? Or Assembly of the fridge as a whole?

As an aside I 112 2nd your hypo for being amicable to two separate interpretations.

Either way, you're tossing in method step limitations into a product claim to make a point and I won't tricked into such tomfoolery. The method step limitations in the actual claim are bad enough, and would have killed a claim before me outright. This claim is being supported, I presume, on the magic of a presumption of validity even though the court didn't use any magic words.

Also as an aside, the claim was ruled to be valid by a court of law, and thus, if I were to ever have an opinion on the validity of this claim (which I certainly never would) then it would have to be in line with the court's ruling.

"a relatively resilient end edge portion which temporarily deflects and subsequently rebounds to snap-secure one of said glass piece front and rear edges in the glass piece edge-receiving channel of said at least one front and rear frame portion,""

As it stands, the claim is a product claim. The product must be able to perform this "method step"/functional recitation in order to snap secure with another object, at any time which we choose to view it, presumptively, and I'm not arguing otherwise. That is, at any given moment in time, we can reach into time and pull out this product (after the edge is finished being formed) from the flow of time. That product which we plucked out of time and is now "timeless" (that is it doesn't change) better meet every limitation of the claim.

They apparently picked the product out of time whilst it was in Mexico. And doubtless, if mexico were a part of the US then the item would infringe. If we pull the item out of the flow of time when it crossed the border it no longer infringes. This is why nirpa argues the court used some provision in the law that allows it to be infringment if the product was made by a process in a foreign country and imported.

"(g) Whoever without authority imports into the United States or offers to sell, sells, or uses within the United States a product which is made by a process patented in the United States shall be liable as an infringer"

He takes liberties with the meaning of this passage. On its face it provides for products made by a process, where the process by itself is patented, to be infringing. There isn't even a process claim at issue here.

"you would have required the Applicant to amend the claim to specify a limiation to being able to deflect and rebound post assembly. "

I would have required him to get that method step limitation the frak out of the claim is what I would have done. And then, when he put a functional limitation in to replace it then I would probably allege that my prior art was capable of performing the function if it was. And then if he wanted to narrow his functional language down some, we'd have to see about that. Maybe I'd just use nirpa's definition of deflection and rebounding and my prior art would read all over that limitation.

Posted by: 6000 at July 27, 2009 12:29 PM

And yes, I'm a bit argumentative, I hate to see sorcery in the courtroom.

I hate it.

Hate it.

Hate.

On the other hand, SG didn't do a good job with its arguments.

Posted by: 6000 at July 27, 2009 12:31 PM