ZeTec Inc v. Westinghouse Electric Co LLC

1. Case Identification

• Case #: To Be Assigned
• Patent #: 6,823,269
• Filed: January 28, 2014
• Petitioner(s): Zetec, Inc.
• Patent Owner(s): Westinghouse Electric Co., LLC
• Challenged Claims: 1-18

2. Patent Overview

• Title: Eddy Current Data Union
• Brief Description: The ’269 patent is directed to methods for synthesizing non-destructive examination (NDE) data for training analysts or testing equipment. The method involves combining noisy NDE data collected from a component at a field site with NDE data generated from a laboratory specimen that simulates the component and includes a selected flaw, creating a "combined data train" that reflects a realistic examination response.

3. Grounds for Unpatentability

Ground 1: Anticipation of Claims 1-10 (wherein claims 1 and 9 are anticipated) by Sullivan

• Prior Art Relied Upon: Sullivan (“Guidelines for Qualifying Eddy Current Technology for CANDU Steam Tube Inspection,” a 1999 report, Ex. 1002).
• Core Argument for this Ground:
  • Prior Art Mapping: Petitioner argued that Sullivan disclosed every element of the challenged claims. Sullivan described a method called "signal injection" for performance demonstrations, which "can superimpose flaw signals upon signals in data files from eddy current scans of in-service tubes." This involved creating flawed tubes in a laboratory, scanning them to get flaw signals, and superimposing those signals onto data from in-service (field) tubes to simulate flaw distortion from background noise. This process directly maps to the claimed method of combining lab specimen data with noisy field site data.
  • Key Aspects: Petitioner contended that Sullivan’s description of "signal injection" to create combined data signals for qualifying inspection techniques is functionally identical to the ’269 patent’s method of synthesizing data for training analysts.

Ground 2: Anticipation of Claims 1-10 (wherein claims 1-9 are anticipated) by Begley et al.

• Prior Art Relied Upon: Begley et al. (“Depth Based Structural Analysis Methods of SG Circumferential Indications,” a 1997 report, Ex. 1018).
• Core Argument for this Ground:
  • Prior Art Mapping: Petitioner asserted that Begley et al., a report co-authored by the named inventors of the ’269 patent, fully anticipated the invention. The report detailed the creation of a large database for analyst training, which involved producing laboratory samples with flaws representative of operational conditions. It described a process where "sectors of flaw signals" were "cut from the flaw signals of pulled tubes [field data]" and "resynthesized as flaw and NDD [no discernible deterioration, i.e. just noise] signals." This process of editing and inserting flaw data into field data was argued to be a direct anticipation of combining lab and field data to form a training data set.
  • Key Aspects: Petitioner highlighted that Begley et al. disclosed calibrating both field and lab data using an ASME standard flaw and using different types of eddy current detectors for field and lab sites, directly anticipating limitations in dependent claims 2, 3, 4, and 5.

Ground 3: Anticipation of Claims 1-10 (wherein claims 1, 2, and 4 are anticipated) by Hölzl

• Prior Art Relied Upon: Hölzl (6,566,871).
• Core Argument for this Ground:
  • Prior Art Mapping: Petitioner argued Hölzl anticipated the core method of the ’269 patent. Hölzl disclosed a process for testing a workpiece by combining a "pattern signal" obtained from a "reference test specimen with a defined fault" with a "noisy measurement signal" obtained from a component in the field. The signals are cross-correlated to produce a combined signal. Petitioner asserted this directly teaches generating data from a flawed lab specimen and combining it with noisy field data. Hölzl’s disclosure that fault signals are displayed as a function of amplitude and phase was argued to inherently teach the calibration steps recited in claims 2 and 4.
  • Key Aspects: The argument focused on Hölzl's direct teaching of combining a reference signal from a specimen with a defined fault with a noisy field signal, which mirrors the fundamental steps of claim 1.

Ground 4: Obviousness of Claim 1 over Hedengren et al. in view of Hölzl, Junker et al., Begley et al., or Sullivan

• Prior Art Relied Upon: Hedengren et al. (5,371,462), Hölzl (6,566,871), Junker et al. (4,763,274), Begley et al. (report, Ex. 1018), or Sullivan (report, Ex. 1002).

• Core Argument for this Ground:

  • Prior Art Mapping: Hedengren et al. taught an eddy current inspection method involving the subtraction of a processed reference trace from a processed test trace (or vice versa) to produce a combined data set emphasizing flaw signals. The reference trace was from a "clean slot known to be free from defects." Petitioner argued that Hedengren et al. taught all elements of claim 1 except for creating a specimen with a selected flaw.
  • Motivation to Combine: Petitioner contended that a person of ordinary skill in the art (POSITA) would have been motivated to modify the method of Hedengren et al. by using a reference specimen with a known flaw, as taught by Hölzl, Junker et al., Begley et al., and Sullivan. This modification would be an obvious substitution to produce predictable results, specifically to create a more realistic test signal for training purposes by injecting a known flaw signal into a noisy background, rather than just subtracting background noise to find an unknown flaw.
  • Expectation of Success: A POSITA would have had a high expectation of success because combining signals was a well-known technique, and the secondary references all demonstrated the utility of using specimens with known flaws for calibration and testing.

• Additional Grounds: Petitioner asserted numerous additional obviousness challenges against all claims (1-18) based on various other combinations of prior art. These included combinations starting with Holt (4,194,149) or Eberhard et al. (4,920,491) in view of references teaching the use of flawed specimens (like Junker et al., Sullivan, or Begley et al.). The underlying theory for these grounds was similarly based on the motivation to combine known NDE signal processing techniques to create realistic training data.

4. Relief Requested

• Petitioner requests institution of IPR and cancellation of claims 1-18 of the ’269 patent as unpatentable.