Ex parte Ryan, Jr.
Appeal 2009004014; App. No. 09/777,592 ; Tech. Center 3600
Decided: September 25, 2009
The application on appeal was directed to a mail processing system. A representative claim on appeal read:
2. A mail piece verification system for processing mail pieces, the mail pieces having associated therewith respective mail piece data, the system comprising:
a data center in operative communication with a plurality of mail processing centers,
the data center including a plurality of account files corresponding to a plurality of postage metering systems;
the data center being adapted to
store reset data in each of the plurality of account files representative of reset activity associated with the plurality of postage metering systems, respectively;
receive respective mail piece data corresponding to the mail pieces from the plurality of mail processing centers;
store empirical data in each of the plurality of account files representative of mailing activity associated with the plurality of postage metering systems, respectively;
conduct a forensic accounting analysis of the empirical data and the reset data associated with a selected postage metering system using a previously defined time period over which to conduct the forensic accounting analysis; and
download graphic data to the selected postage metering system to be included in the mail piece data of mail pieces subsequently prepared by the selected postage metering system if the forensic accounting analysis reveals that the empirical data is not consistent with the reset data for the selected postage metering system.
The Examiner rejected claim 2 as obvious over Hunter, Moore, and Connell. The issue on appeal was whether the combination of Hunter and Connell taught the elements emphasized above.
Specifically, in the Final Office Action the Examiner relied on Hunter for the "conduct a forensic accounting analysis" action, for the claimed "if" condition, and for taking an action – further investigation of possible fraud – in response to the claimed "if" condition. Since Hunter's responsive action was not the claimed "download graphic data" action, the Examiner then relied on Connell as teaching the "download graphic data" action in response to detection of fraud. In explaining how and why these features were combined, the Examiner used the following reasoning:
To paraphrase in order to explain the rejection in a more simple manner, Hunter detects potential fraud by keeping track of postage value added to a meter, as well as mail pieces from that meter that pass through the postal system. If the postage having been purchased for a meter differs too much from the amount of postage on mail pieces from that meter that have been mailed, this is an indication of potential postage fraud. When such fraud is detected, Hunter initiates a responsive action to deal with the situation. In Hunter, that responsive action is to initiate an investigation (see Hunter, column 5, lines 31-39).
Of course, claim 1 requires that the responsive action be not an investigation, but a download of graphic data into a meter that gets incorporated onto mail pieces for the purpose of identifying mail pieces originating from suspect meter users. It is necessary to turn to Connell et al. for this portion of the claim limitation.
Connell et al. is a postal security system that also functions in response to possible postal fraud (see Connell et al., column 1, lines 30-52). Connell et al. works by incorporating graphic data onto mail pieces that indicates whether or not the postage on the mail pieces is legitimate. At any given point in time, there is a particular item of graphic data that is designated as indicating that a mail piece's postage is valid. This designated graphic data changes periodically. In order to keep their "designated graphic data" up-to-date so that their mail pieces are indicated as having valid postage, meter users in Connell et al. must occasionally download the latest version of the designated graphic data. Individuals who are not authorized to be metering mail pieces will not have access to up-to-date designated graphic data, and thus will mail mail pieces with out-of-date graphic data that indicate the mail pieces as potentially fraudulent. Therefore, Connell et al. provides the needed disclosure of graphic data on mail pieces that is used to identify potentially fraudulent mail pieces in response to meter fraud concerns.
Since the out-of-date graphic data used by a potentially fraudulent user was at one point up-to-date graphic data that was downloaded validly to that user, the out..:of-date graphic
data in Connell et al. that identifies potentially fraudulent data is disclosed as being downloaded
to user meters.
On appeal, the Applicant did not argue the teachings of the individual references, but instead challenged the manner in which the features were combined. Specifically, the Applicant argued that the trigger for Connell's download action was much different from the trigger for Hunter's further investigation action:
[T]he system in Connell selects one of a plurality of stored indicia images [for download] based solely on the date that the indicia will be printed. ... The date has nothing whatsoever to do with whether or not a forensic accounting analysis reveals that the empirical data is not consistent with the reset data. ... Without using the present claims as a road map, it would not have been obvious to make the multiple, selective modifications needed to arrive at the claimed invention from these references.
In the Answer, the Examiner maintained the rejection and also provided additional explanation.
The use of graphic data to identify postal fraud, as can be seen in the above passages from Connell et al. is a responsive action to the overall problem of postage indicium fraud. While the responsive action in Connell et al. is not a responsive action in the same specific and immediate way that the responsive action in Hunter is, the responsive action in Connell et al. is still a responsive action.
Also, while the Answer continued to rely on a specific motivation in Connell, the Answer brought in one of the KSR rationales, asserting that the claims “only unite old elements with no change in their respective functions and which yield a predictable result.”
The Board reversed the obviousness rejection. Noting the Examiner's reliance on the "predictable results" rationale from KSR, the Board instead found that the combination used by the Examiner was "more than the predictable use of a prior art element according to its established function." The Board explained this statement as follows:
The Examiner states, “. . . the responsive action in Connell et al. is not a responsive action in the same specific and immediate way that the responsive action in Hunter is . . ..” Answer 26. We agree. While the overall invention in Connell may be directed to the prevention of postage indicia fraud, Connell describes the “responsive action” (i.e. the downloading of an address of a newly authorized indicia) relied upon by the Examiner functions to detect the postage indium fraud. The Examiner’s “responsive action” of Connell is not a response to a detection of an occurrence of postage indicia fraud but is a means that functions to detect an occurrence of the fraud. Connell’s “responsive action” performs the same function as the Hunter’s description of comparing purchased postage with used postage and not the further investigation of Hunter, as the Examiner seems to assert (see Answer 23). Further, Connell’s “responsive action” performs the same function as the claimed forensic analysis and not the claimed downloading of graphic data. ... Therefore, we find that the Appellant has shown the Examiner erred in rejecting claim 2.
My two cents: First, I give credit to the Examiner for providing a detailed explanation of the rejection. Second, I'm glad to see a successful attack on the common "predictable results" rationale. However, I think this fact pattern better fits a proposed-combination-does-not-teach-the-claim argument. That said, I think I figured out how to make the fact pattern fit the predictable results rationale. (Or, more precisely, the UNpredictable results rationale.)
The way I see it, the Examiner's combination rejection takes the form
- Claim includes action A (forensic analysis), result B (data not consistent) of action A, and action C (downloading graphic data) in response to result B.
- Hunter teaches action A, result B of action A, and action C' (initiate fraud investigation) in response to result B.
- Connell teaches action C in response to result B.
- Combination results in action A, result B of action A, and action C.
The Board disagreed with the Examiner's reading of the reference. That is, the Board found that Connell's downloading graphics data was not properly understood as action C in response to result B, but was instead action D (fraud detection). Thus, the Board found that the Examiner's combination did not produce the claim, which required action A, result B of action A, and action C in response to result B.
That's why I see the Board's reasoning as a natural fit for a proposed-combination-does-not-teach-the-claims argument, rather than an unpredictable-results argument.
But after thinking about this decision for a while, I think I see how to map this very same reasoning to the lack-of-predictable-results rationale. You could say that the predictable combination of Hunter and Connell is A, B as a result of A, C' in response to B, and D. But the claims require A, B as a result of A, and C in response to B. So the claim is not a predictable result of the combination, and the Examiner didn't make a prima facie case of obviousness.
As a side note, you'll recall that the Examiner mentioned the predictable results rationale in the Answer, but also used an explicit motivation to combine from the secondary reference. The Board didn't address this alternative rationale.
Finally, I note that the Applicant also used a hindsight argument during prosecution and in the Appeal Brief, and the Board also found hindsight. Personally, I find hindsight to be a particularly UNhelpful way of viewing obviousness. Hindsight is the conclusion reached when the prima facie case isn't made ... the Examiner must have used hindsight because he hasn't made a prima facie case. Hindsight is not itself a standalone argument against obviousness.