Thursday, May 22, 2014

Federal Circuit rules on indefiniteness standard and on "adapted to"

The recent Federal Circuit decision In re Packard looked at the standard for indefiniteness used by the PTO. During prosecution, the PTO applies the "two or more plausible interpretations" standard, per the precedential Board decision Ex parte Miyazaki. The standard applied by the courts for an issued patent is instead "insolubly ambiguous." The court in In re Packard declined to reach the larger question of whether the PTO can use this different standard, and decided the narrower question of whether the indefiniteness rejection was proper. The court affirmed the rejection because the Applicant (pro se inventor) didn't substantively respond to the indefiniteness rejection.

In re Giannelli from earlier this year was also a review of a PTAB decision, this one involving claim construction of "adapted to." The PTAB affirmed an obviousness rejection of a claim to an exercise machine, taking the position that the prior art chest press was "capable of" performing the function recited in the claim. The Federal Circuit reversed, after finding that in this particular case, "adapted to" required more than mere capability. The court relied on the specification's disclosure of embodiments that were specifically designed for rowing.

Ryan Alley's blog has thoughtful coverage of both decisions:
   In re Packard – Claim Drafting and Prosecution Lessons
   In re Giannelli – Claim Drafting and Prosecution Lessons

Ryan's blog is my favorite Federal Circuit blog because it offers many practical lessons for prosecutors. 

Tuesday, May 20, 2014

When does "a" mean "the"?

Claims use "a" and "the" in ways that are at odds with plain English usage, which can lead to some strange results.

Let's say I'm describing a text processing algorithm which counts how many times a pattern is found. In plain English, it's perfectly natural to express this as "THE number of occurrences". However, claim drafting reserves "the" for previously-introduced elements. So a claim would probably use "a" instead, to read "A number of occurrences".

Note that in plain English this construction has a completely different meaning: "a number of" often means "some" or "several" (e.g.,"a number of widgets were defective"). Even so, usually no one makes an issue of using "A" to mean "THE" even though it's counter to the plain English meaning. Our brains have been trained to apply a sort of patent filter when reading claims, and we intuitively understand that the drafter really meant "THE number of occurrences even though the claim reads "A number of occurrences."

But what if you don't assume the drafter was using patentese rather than plain English? What if you instead take the position that "a number of" has two different meanings. Then isn't the claim indefinite?

In the recent decision Ex parte Whitney (PTAB 2013), the Board said Yes, it is indefinite. The claim language at issue was: "comparing a number of occurrences of a text pattern in the collection of text at one moment in time with numbers of occurrences of the text pattern in the collection of text at other moments in time.” The Board explained as follows:
     Applicant's arguments ... focus on the failure of the prior art to disclose or suggest comparing an actual numerical count of occurrences, i.e., “a number of occurrences,” of a text pattern in a collection of text at one moment in time with the actual numerical count of occurrences of the text pattern in the collection of text at other moments in time. However, it is equally proper to interpret the phrase, “a number of occurrences,” as the Examiner does, to refer to an indefinite group of occurrences of a text pattern, e.g., some occurrences or several occurrences of the text pattern that are compared at different points in time based on some unspecified criteria.
     Because the claim language is susceptible to two equally reasonable interpretations, the claim language is ambiguous, and, thus, indefinite. See Ex parte Miyazaki, 89 USPQ2d 1207, 1215 (BPAI 2008) (precedential). Therefore, we will enter a new ground of rejection of claims 1-50, 53, and 54 as indefinite under 35 U.S.C. § 112, second paragraph.
The Examiner did not raise this issue. And because the claim was indefinite, the Board refused to decide on the prior art rejection. (In re Steele, 305 F.2d 859, 862 (CCPA 1962)(A prior art rejection cannot be sustained if the hypothetical person of ordinary skill in the art would have to make speculative assumptions concerning the meaning of claim language.)

A few months earlier, the Board reached the opposite result in Ex parte Qu (PTAB 2013). The limitation at issue read "wherein an arithmetic sum or difference of twice a number of stator teeth and a number of the stator poles equals a number of rotor poles." The Examiner asserted that it was unclear whether "a number of" ... "intended to refer to the total number of teeth or poles or some subset thereof.” The Examiner then interpreted the phrase to mean "subset of teeth" in the prior art rejection. The Examiner explained as follows: 
[T]he devices disclosed in Qu satisfy equation (1) because any number of rotor poles, stator poles, and stator teeth can meet the claimed limitation since ‘a number’ can be any integer, up to the maximum number of poles/teeth disclosed by Qu.
The Board found the claim was not indefinite in view of the specification:
In determining whether a particular claim is definite, the claim’s language must be analyzed in light of the content of the particular application, the prior art’s teaching, and the interpretation that would be given to the claim’s language by a person of ordinary skill in the art at the time the invention was made. In re Skvorecz, 580 F.3d 1262, 1268 (Fed. Cir. 2009).
The Board found that an equation in the specification made clear that "A number of" meant "THE number of":
The number of the rotor poles, the number stator poles and the stator teeth are configured to satisfy: 2 * Steeth ± Spole = Rpole, wherein STEETH refers to the number of stator teeth 30, SPOLE refers to the number of stator poles and RPOLE refers to the number of rotor poles.
Using this construction, the Board then reversed the prior art rejection.

I think these two decisions are outliers. In the first place, I think most Examiners use the same "patentese" rules of antecedent basis that claim drafters do. Most don't reject as indefinite, and most don't take the position that "A number of" means "SOME of". So the Board probably doesn't see this issue raised by the parties very often. I also looked for other decisions where the Board raised the issue, and didn't see any.

But these decisions do highlight how very different claim language is from plain English.

Wednesday, May 14, 2014

Is it circular reasoning when an obviousness rejection says the reason for adding a feature is to get that feature?

Takeaway: Sometimes the Examiner asserts that the reason to add a feature from a secondary reference is to obtain that very same feature. Is this circular reasoning and thus an improper rationale for obviousness? Today's post highlights a few cases where the PTAB said this logic was insufficient to support a conclusion of obviousness.

Discussion: In many obviousness rejections, the Examiner's rationale for combining amounts to "add feature X from reference B to reference A in order to get feature X." Yet obtaining the added feature X is only a reason to combine if having X provides an undisputed advantage, right? Otherwise, it's If the mere presence of X doesn't provide such an advantage, then the Examiner still hasn't shown why a POSITA would combine A and B.

In today's post, I'll discuss a few cases where the Examiner used this sort of reasoning and the Board reversed. Now, I'm not saying that this argument is a winner. In many of the decisions that I review, I end up convinced that X does convey an undisputed advantage, and thus does serve as a proper reason to combine. But if you have a case where you think "add X to get X" doesn't adequately motivate a POSITA to combine the references, you may find the Board's reasoning in these three decisions to be helpful in framing your argument.

The claim in Ex parte Banna (PTAB 2014) involved a multi-stage receiver including a delay controller. The claim limitation at issue required the controller to "select the delay value for the delay block by calculating a processing delay associated with the generation of the one or more processing parameters." The Applicant appealed an obviousness rejection that used a single reference. The Examiner acknowledged that the single reference (Sendyk) did not teach calculating a processing delay, but asserted that it would have been obvious “to modify the invention of Sendyk, and to obtain the time difference between the two delay control signals to have the delay value.” (Emphasis added.) The Applicant argued that “the Examiner has failed to identify any valid reason to modify Sendyk.” (Emphasis in original.)

The Board agreed that the Examiner did not provide a reason with rational underpinning, and reversed the rejection. The Board cited In re Chaganti, 2014, WL 274514, *4 (Fed. Cir. 2014) (“It is not enough to say that . . . to do so would ‘have been obvious to one of ordinary skill.’ Such circular reasoning is not sufficient–more is needed to sustain an obviousness rejection.")

In Ex parte Foxenland (PTAB 2013), the claim at issue related to ordering a call list of party identifiers (i.e., phone numbers) based on call frequency, where two or more communication identifiers correspond to a single party identifier. More specifically, the claim required that "the call frequency information for the single party identifier includes the call activity related to each of the at least two of the plurality of communication identifiers." The Examiner relied on Kreitzer's arrangement of a call list in order of call frequency based on communication identifiers, and Levy's use of multiple phone numbers associated with a single party. As a reason for incorporating Levy into Kreitzer, the Examiner offered "for the purpose of managing a plurality of communication identifiers that correspond to a single party identifier by storing these data in the communication terminal."

The Board reversed the obviousness rejection. The Board found that the Examiner had concluded obviousness without giving a "specific reason why the ordinary artisan would have had any incentive to incorporate multiple different numbers associated with a single individual as taught by Levy into the hotlist of Kreitzer." (Emphasis added.) Thus, although the Examiner had shown that each claimed feature was known, he did not provide a sufficient reason for combining them. 

The reexamination appeal LG Electronics v Whirpool (PTAB 2013) involved ice dispensing in a bottom mount refrigerator, which purportedly "presented a unique challenge because the freezer compartment is located lower than desired for an ice dispensing location." The Board framed the obviousness issue as whether the Examiner provided a sufficient reason to "incorporate both an ice dispenser and a separate door in the same ice compartment" as required by the claim. 

The Examiner combined a primary reference showing a door providing access into a refrigerator ice making/storage compartment with a secondary reference showing a refrigerator ice dispenser. The rejection used the "predictable result" rationale, asserting that "such a combination would apply known features, such as an ice compartment door, to achieve a predictable result of providing a user access to the ice in the ice compartment." The Board reversed. The Board assumed that "such a combination would apply known features to achieve a predictable result" – but this is true only after the decision is made to combine these features. The rejection failed because the Examiner did not explain "why one would have chosen to combine both a prior art access door and a dispenser in a single refrigerator in the first place." (Emphasis in original.)

Thursday, May 8, 2014

Board finds that "band of frequencies" does not require range and reads on single clock frequency

Takeaway: The Applicant appealed claims to a memory controller using two frequency bands to communicate with a memory device. The Examiner rejected as anticipated by a reference that taught the use of two clock frequencies (X and X/2). The Applicant appealed and argued that "band of frequencies" implies a range, and thus does not read on a single frequency. The Board affirmed the rejection because the Applicant's specification discussed an embodiment using a "zero band of frequencies", which the Board interpreted as a single frequency.


Ex parte Alon
Appeal 2010008694; Appl. No. 11/021,514; Tech. Center 2600
Decided: Feb. 19, 2013

The application was directed to a memory controller with a simultaneous bi-directional link to the memory. A representative claim on appeal read:

     1. A memory system, comprising:
     a controller;
     a memory device; and
     a set of signal lines, coupled to the controller and the memory device,
     wherein, on each of the signal lines, the controller is to communicate to the memory device using a first band of frequencies while the memory device is communicating to the controller using a second band of frequencies, and
     wherein the controller is configured to dynamically adjust the first band of frequencies based on a predetermined data rate from the controller to the memory device
and to dynamically adjust the second band of frequencies based on a predetermined data rate from the memory device to the controller.

The Examiner asserted that Koroodi's teaching of a clock rate of X Mhz and a clock rate of X/2 MHz read on the "first band of frequencies" and "second band of frequencies". The Applicant appealed and argued that reading "band of frequencies" on a single clock frequency was an unreasonably broad interpretation of "band".

The Applicant made several points related to Broadest Reasonable Interpretation. According to the Applicant, "the clock rate of a controller is a distinct technical concept from frequency bands for communication, such as for communication from a controller to a memory device or a memory device to a controller", and the Examiner violated the BRI requirement by equating these "distinct concepts".

Moreover, the Applicant argued, "band of frequencies" means a range of frequencies, and does not read on a single frequency as asserted by the Examiner. As support for this argument, the Applicant referred to Fig. 3's depiction of "first band of frequencies 364" as encompassing a range of frequencies 368. According the Applicant, this would teach a person of ordinary skill in the art (POSITA) that "band of frequencies" included multiple frequencies. The Applicant then referenced other uses of the term "band" in the art, asserting that "a band-pass filter passes a specified range of frequencies while filtering out frequencies outside the specified range." According to the Applicant, this shows that a POSITA would understand "band" as "range", even without consulting the specification.

Neither party disputed that Kuroodi taught a single clock frequency rather than a range. Thus, the Applicant concluded that under a reasonable interpretation, Kuroodi did not teach the claimed "range of frequencies."

In the Answer, the Examiner responded to the Appeal Brief arguments by stating that "the Examiner is not limited to Applicant's definition, which is not specifically set forth in the claims. In re Tanaka et al. 193 USPQ 139, (CCPA) 1977.

The Board found the Examiner's interpretation to be reasonable, and affirmed the anticipation rejection. After explicitly adopting the Examiner's finding and reasoning, and concurring with the Examiner's conclusions, the Board offered this additional explanation.

The Board first indicated that the Examiner's requirement for a definition to be incorporated into the claims was improper. The Board then noted that the claim term should be afforded its plain meaning since the Applicant's specification did not provide a definition for the term.

The Board was not persuaded by the Applicant's first BRI argument – that clock rate and communication frequency – are different, and instead concluded that a clock rate may be expressed as a frequency.

As for the second BRI argument  – that "band" did not encompass a single frequency – the Board found that the specification taught to the contrary. More specifically,  para. 044 of the Applicant's specification "provides an example in which there is a 'zero band of frequencies' ". ( "The memory system 200 may have a mode of operation in which the control logic 210 allocates a substantially zero band of frequencies to the first band of frequencies 368 (FIG. 3) and a maximum band of frequencies to the second band of frequencies 370 (FIG. 3) ..."

The Applicant made other arguments as well, but the Board was not persuaded by these either, and affirmed the anticipation rejection.

My two cents: The Board interpreted "zero band" as a single frequency. But it seems to me that the discussion of "zero band" in Applicant's spec was really saying that one embodiment allocated NO frequencies to one direction, and all of the frequencies to the other. In fact, I'd say this is how a POSITA would understand the spec as written.

Perhaps if the drafter had used this more straightforward phrasing, instead of "zero band", the Board would have ruled differently. This embodiment was captured in an original dependent claim. Perhaps the drafter was taught to avoid claiming a negative ("allocate no frequencies")? Could the Applicant have covered this embodiment by simply claiming "allocate all of the frequency band" to the other direction?