PTAB
IPR2019-00619
Tetra Tech Canada Inc v. Georgetown Rail Equipment Co
Key Events
Petition
Table of Contents
petition
1. Case Identification
- Case #: IPR2019-00619
- Patent #: 7,616,329
- Petitioner(s): Tetra Tech Canada Inc.
- Patent Owner(s): Georgetown Rail Equipment Company
- Challenged Claims: 16, 23, 38, 44, 46, 62, 68, and 70
2. Patent Overview
- Title: System and Method for Inspecting Railroad Track
- Brief Description: The ’329 patent discloses a vehicle-mounted inspection system that uses a laser and optical receiver (camera) to inspect railroad tracks. The system projects a line of light across the track, captures 2D contour range images of track components, compiles them into a 3D representation, and analyzes the data to detect defects like misaligned or sunken tie plates.
3. Grounds for Unpatentability
Ground 1: Obviousness over Davis, Kowalski, Bostrom, Trosino, and Kanade - Claims 16, 23, 38, 44, 46, 62, 68, and 70
- Prior Art Relied Upon: Davis (a 1990 report on tie performance), Kowalski (Patent 7,023,539), Bostrom (Patent 6,496,254), Trosino (Patent 6,356,299), and Kanade (a 1987 publication on machine vision).
- Core Argument for this Ground:
- Prior Art Mapping: Petitioner argued that Davis taught the fundamental concept of measuring "tie plate cut" depth—the difference in elevation between a tie plate and the adjacent crosstie—as a key indicator of railroad tie deterioration. Kowalski taught a vehicle-mounted, automated laser-scanning system for inspecting railroad tracks and detecting alignment defects by comparing the height contours of track components. The combination of Davis and Kowalski provided the core system. Bostrom was cited for teaching a range-imaging vision system that generates a plurality of 2D contour range images from a scanned object. Trosino taught an automated track inspection vehicle that uses a global positioning system (GPS) for time-stamping and geolocation of detected anomalies. Finally, Kanade was cited for teaching the common machine-vision technique of using an "evaluation zone" or "region of interest" (ROI) to focus analysis on specific parts of an image to accelerate processing and improve accuracy.
- Motivation to Combine: A POSITA would combine Davis and Kowalski to automate the manual or semi-automated process of measuring tie plate cut, thereby improving the speed, objectivity, and safety of railroad inspection. Automating this known inspection task was a well-established goal in the art. A POSITA would further incorporate Bostrom's 2D imaging, Trosino's geolocation, and Kanade's ROI analysis as well-known techniques to enhance the functionality, efficiency, and utility of the automated inspection system, representing predictable improvements.
- Expectation of Success: The combination involved applying known automated machine vision technologies (Kowalski, Bostrom, Kanade, Trosino) to a known railroad inspection metric (Davis's tie plate cut). This integration was considered straightforward and would have yielded the predictable result of an automated system for detecting misaligned or sunken tie plates.
Ground 2: Obviousness over Holmes, Bostrom, Trosino, Kanade, and Davis - Claims 16, 23, 38, 44, 46, 62, 68, and 70
- Prior Art Relied Upon: Holmes (Patent 6,647,891), Bostrom (Patent 6,496,254), Trosino (Patent 6,356,299), Kanade (a 1987 publication), and Davis (a 1990 report).
- Core Argument for this Ground:
- Prior Art Mapping: This ground presented an alternative combination, using Holmes as the primary reference. Petitioner asserted that Holmes taught a vehicle-mounted, range-finding based image processing system for identifying and inspecting railway features like ties and tie plates by analyzing contour information from a generated "range-based virtual surface." While Holmes provided a suitable base system for automated inspection, Petitioner argued it did not explicitly teach the specific algorithm for detecting a sunken tie plate. Davis supplied this missing element by teaching the method of measuring tie plate cut depth by comparing the elevation of the tie plate to the crosstie. The secondary references—Bostrom, Trosino, and Kanade—were argued to provide the same incremental features as in Ground 1: generating plural 2D images, adding geolocation, and using ROI-based analysis, respectively.
- Motivation to Combine: A POSITA would be motivated to implement Davis's specific and reliable tie plate cut algorithm into Holmes's more general railway inspection system. This would provide an additional, objective performance measure of crosstie health, improving the system's ability to identify defects and enhancing overall railroad safety. Adding the teachings of Bostrom, Trosino, and Kanade were again presented as predictable improvements to enhance the system's capabilities.
- Expectation of Success: Petitioner contended that incorporating a known measurement technique (Davis) into a known type of inspection system (Holmes) would have been a simple and predictable modification for a POSITA. The successful integration of these established technologies was reasonably expected.
4. Key Claim Construction Positions
- Petitioner argued for specific constructions of key terms that it contended were critical to the invalidity analysis.
- "frame" and "image": Petitioner proposed construing these terms as "a collection or data structure of pixels forming a two-dimensional representation." This construction was argued to be consistent with the ’329 patent’s specification and figures, which depict a single 2D contour profile as a "frame." Petitioner asserted that a broader construction that could encompass 3D data would be improper, as the patent consistently distinguishes between individual 2D frames and the "compilation" of frames that forms a 3D representation.
5. Arguments Regarding Discretionary Denial
- Petitioner argued that the proposed grounds were not cumulative of issues previously considered by the U.S. Patent and Trademark Office. The petition highlighted that during both the original prosecution and a subsequent ex parte reexamination, the Examiner did not consider Davis, Kowalski, or Bostrom, nor the specific combinations asserted in the petition. Petitioner contended that these new combinations and references teach the very features the Examiner had previously found lacking in other prior art, namely the use of machine vision and specific processing algorithms to detect a misaligned or sunken tie plate.
6. Relief Requested
- Petitioner requested the institution of an inter partes review and a final written decision finding claims 16, 23, 38, 44, 46, 62, 68, and 70 of the ’329 patent unpatentable, followed by their cancellation.
Analysis metadata