Regarding Invalidity of U.S. Patent No. 8,095,237 Peter I. Corke, Visual Control of Robots: High-Performance Visual Servoing (Aug. 1996) (“Corke”) Guo-Qing Wei, et al., “Active Self-calibration of Robotic Eyes and Hand-eye Relationships with Model Identification,” TRANSACTIONS ON ROBOTICS AND ANIMATION, vol.

14, No. 1, Feb 1998 (“Wei-I”) “Multisensory Visual Servoing by a Neural Network,” Guo-Qing Wei et al., IEEE Transactions on Systems, Man and Cybernetics, April 1999 (“Wei-II”).

1007 RoboticVISIONTech, Inc. v. ABB Inc., No. 1:22-cv-01257, Summons in a Civil Action (served Sept. 23, 2023).

1008 U.S. Patent No. 4,146,924 to Birk et al. 1009 U.S. Patent No. 5,959,425 to Bieman et al. 1010 Declaration of Sylvia Hall-Ellis, Ph.D. 1011 Docket Sheet, RoboticVISIONTech, Inc. v. ABB Inc., Case No. 1:22-cv-01257-GBW (D.

Exhibit Description 1012 Curriculum Vitae of Seth Hutchinson, Ph.D. 1013 Scheduling Order, RoboticVISIONTech, Inc. v. ABB Inc., Case No. 1:22-cv-01257-GBW (D.

The patent explains that prior “camera systems [] have used laser triangulation which involves expensive specialized sensors, must be rigidly packaged to maintain geometric relationships, require sophisticated inter-tool calibration methods and tend to be susceptible to damage or misalignment when operating in industrial environments.” Id. at 1:32–37.

However, it is blackletter law that when an element is not taught by the prior art, “there must be a showing of a suggestion or motivation to modify the teachings of that reference to the claimed invention in order to support the obviousness conclusion.” SIBIA Neurosciences, Inc. v. Cadus Pharm. Corp., 225 F.3d 1349, 1356 (Fed. Cir. 2000).

For instance, Corke provides an example of “visual servoing” in which “[t]hree robot DOF [(degrees of freedom)] are controlled by image features so as to keep the camera at a constant height vertically above a target rotating on a turntable,” as illustrated in Corke’s Figure

As explained below, Petitioner’s argument makes no sense temporally or logically because the claimed transformation needs to be based at least in part on features located in the “target object,” but the “target object” is not present in Corke when the camera is being calibrated; it comes into play only afterwards.

Specifically, the parties have exchanged initial infringement and invalidity contentions, document production and interrogatory responses are well underway, and all claim construction briefing will be completed prior to April 4, 2024, the date by which an institution decision is due here.

Mail Stop “PATENT BOARD” Patent Trial and Appeal Board U.S. Patent and Trademark Office P.O.

These Notices are being timely filed within 21 days of service of the Petition.

The Patent Owner and real party-in-interest is RoboticVISIONTech, Inc.

U.S. Patent No. 8,095,237 (“the ’237 patent”) is at issue in the pending district court proceeding RoboticVISIONTech, Inc. v. ABB Inc., Case No. 1:22-cv- 01257-GBW (D.

Please address all correspondence to the lead and back-up counsel as shown above.

Mail Stop “PATENT BOARD” Patent Trial and Appeal Board U.S. Patent and Trademark Office P.O.

Box 1450 Alexandria, VA 22313-1450

Case IPR2023-01426 U.S. Patent No. 8,095,237 Pursuant to 37 C.F.R. § 42.10(b), Patent Owner, RoboticVISIONTech, Inc., hereby appoints the practitioners associated with Customer Number 26111 and all other practitioners listed below asits attorneys to transact all business in the United States Patent & Trademark Office associated with the above-captioned inter partes review of U.S. Patent No. 8,095,237.

These attorneys includeat least: Daniel S. Block (Reg. No. 68,395) Lead Counsel: Back-up Counsel: Steven M. Pappas (Reg. No. 73,904) Address:

Case IPR2023-01426 U.S. Patent No. 8,095,237 CERTIFICATE OF SERVICE (37 C.F.R. § 42.6(e)) I certify that the above-captioned PATENT OWNER’S POWER OF ATTORNEY was served in its entirety on October 9, 2023, upon the following parties via electronic mail:

Petitioner, ABB, Inc., hereby appoints the following practitioners as ABB, Inc.’s counsel to transact all business in the United States Patent and Trademark Office in connection with the above-captioned inter partes review.

Dated: ACTIVE 690358305v1 September 20, 2023

Placing the camera on the robot’s arm—corresponding to the claimed moveable portion of a robot—provides certain advantages, including “the ability to avoid occlusion, resolve ambiguity and increase accuracy, by directing its attention.” EX1004, pp.3, 166.

Corke explains that four images of the SHAPE calibration target were used to compute the “calibration matrix”—a mathematical object that “contains information about the position of the camera with respect to the world coordinate frame”—the extrinsic parameters.

Because a “training space” is a “reference frame defined with respect to a point on the calibration” object, a POSITA would have understood that Corke discloses extrinsic parameters, including a camera space-to-training space transformation.

A POSITA would have understood that Corke teaches an algorithm that employs “features extracted from the image” (only the single captured image) and “a geometric model of the target” (known physical relationship) to determine the “3D relative pose of an object” (object space-to-camera space transformation for the

In fact, Patent Owner overcame Wei-II on the basis that “Wei[-II] …is directed to the use of a stereo pair of cameras” but the claimed invention is “directed to methods and apparatus that employ single camera three-dimensional (3-D) vision for robotic guidance.” EX1002, p.129.

Exhibit 5 Automaker Study Finds eVisionFactory Most Accurate 3D Guided Robotics Vision System Silver Spring, Md.

“Determining all six degrees of freedom accurately is one of the greatest challenges in the field of vision guidance for robotics,” said RVT VP, engineering Paul Weidinger.

“Our platform enhanced accuracy beyond either the camera resolution or the robot itself, showing the critical nature of software in the success of robotics in all fields.” The study included a series of tests to gage the repeatability and reproducibility (gage R&R) of positional calculations, an industry standard method that objectively compares the accuracy between vision systems.

Outlined in the chart below, RVT’s two tests displayed a higher accuracy than every other vision system studied in every category: Max Deviation Requirements ±3.0mm

RVT’s eVF™ software platform is installed in hundreds of systems worldwide and operates every day controlling over one-half billion dollars of capital equipment.

09/19) The JS 44 civil cover sheet and the information contained herein neither replace nor supplement the filing and service of pleadings or other papers as required by law, except as provided by local rules of court.

This form, approved by the Judicial Conference of the United States in September 1974, is required for the use of the Clerk of Court for the purpose of initiating the civil docket sheet.

Habeas Corpus: ’ 791 Employee Retirement ’ 440 Other Civil Rights ’ 463 Alien Detainee Income Security Act ’ 441 Voting ’ 510 Motions to Vacate ’ 442 Employment Sentence ’ 443 Housing/ ’ 530 General Accommodations ’ 445 Amer.

’ 462 Naturalization Application ’ 465 Other Immigration Actions V. ORIGIN (Place an “X” in One Box Only) ’ 1 Original ’ 2 Removed from Proceeding State Court ’ 3 Remanded from Appellate Court ’ 4 Reinstated or Reopened ’ 6 Multidistrict Litigation - Transfer ’ 8 Multidistrict Litigation - Direct File ’ 5 Transferred from Another District (specify) Cite the U.S. Civil Statute under which you are filing (Do not cite jurisdictional statutes unless diversity): Brief description of cause:

This form, approved by the Judicial Conference of the United States in September 1974, is required for the use of the Clerk of Court for the purpose of initiating the civil docket sheet.

A figure in page 171 of the ABB User Manual (reproduced below) shows software images from two cameras in which 9 features are visible on an object.

A figure in page 171 of the ABB User Manual (reproduced below) shows software images from two cameras in which 9 features are visible on an object.

Figures from pages 180, 181, and 182 (respectively) showing 9 features identified on the part that have an invariant physical relationship to one another based on various coordinate systems.

“Temperature: 0°C to 50°C (w/HDD), Extended Temperature: -20°C to 55°C (-4 to 131°F) w/industrial SSD or CFast…The Processing Unit used for the current version of FlexVision™ is Matrix MXC-6300 Series from ADLINK Technology Inc.” ABB User Manual, p. 9.

A figure in page 171 of the ABB User Manual (reproduced below) shows software images from two cameras in which 9 features are visible on an object.

Except as may be expressly stated anywhere in this manual, nothing herein shall be construed as any kind of guarantee or warranty by ABB for losses, damages to persons or property, fitness for a specific purpose or the like.

· · · System Information > Hardware > Cameras Basler acA1600-20gm Appearance Dimensions ABB Part Number: 3HUB502762-1101 Resolution: 1626 px x 1236 px Pixel Size: 4.4 µm x 4.4 µm Interface: GigE Pixel Bit Depth: 12 bits Signal to Noise Ratio: 39,2 dB I/O: 1 input, 1 output Lens Mount: C-Mount Power Requirements: PoE or 12 VDC Power Consumption: 3.4W · Weight: 90g Conformity: CE, RoHS, GenICam, GigEVision, IP30, UL, FCC, IEEE 802.3af (PoE) Sensor: Sony ICX274, CCD (7.16mm x 5.44mm) Shutter: Global Maximum Image Circle: 1/1.8"

18W Strobe Input: NPN > GND (<1VDC) to activate NPN Line: 22mA @ Common (0VDC) Continuous Mode: Light will be in continuous mode by leaving signal on strobe input active Connection: 4 pin 2.5mm pitch phoenix connector Daisy Chain: Up to eight LHF300 Ambient Temperature: -20º - 50º C (-4º - 122º F) Lifespan: 100,000 hrs Color: 625nm Red IP Rating: IP50 · Weight: ~455g

To install PCIe GIE 64+ into Matrix MXC 6300 series, follow the steps below: Withdraw the thumbscrew and remove the top cover.

Figure 7: Network Adapter Camera Port Properties (Cognex GigE Vision Configuration Tool) © Copyright 2016 ABB.

Effective Date of Registration: September 16, 2022 Registration Decision Date: September 16, 2022 This Certificate issued under the seal of the Copyright Office in accordance with title 17, United States Code, attests that registration has been made for the work identified below.

The information on this certificate has been made a part of the Copyright Office records.

»: is ~: 'Ft United States Register of Copyrights and Director

Copyright Office notes: Regarding claimant information: Statement added by Copyright Office from additional information provided by applicant.

Page 2 of 2

A method useful in three-dimensional pose estimation for use with a single camera mounted to a movable portion of a robot, the method comprising: The ABB User Manual discloses detailed structures and functions for handling an object using, e.g., the robot shown at page 108 (reproduced below).

figure in page 171 of the ABB User Manual (reproduced below) shows a software image in which 9 features are visible on an object.

It is the transformation from the Camera Coordinate System to the work frame used in RAPID code.” ABB User Manual, p. 99.

“The Processing Unit used for the current version of FlexVision™ is Matrix MXC-6300 Series from ADLINK Technology Inc.” ABB User Manual, p. 9.

The robot can then properly initiate the next sequence in the tasks given to it and accurately determine where the part being used is.” ABB User Manual, p. 172.

Sanchez, A., et al., "Robot-Arm Pick and Place Behavior Program ming System Using Visual Perception," in Proceedings of the 15th International Conference on Pattern Recognition, Los Alamitos, CA, Sep. 3-7, 2000, 4:507-510.

Drawbacks of such attempts include the need for expen sive specialized sensors as opposed to use of standard off-the-shelf hardware, the need for knowledge of exact geometric relationships between all elements of the system

FIGS. SA and 8B are a flow diagram showing a method of computing three-dimensional positions of the features in the respective coordinate reference frames associated with each image sensor, according to one illustrated embodiment.

A camera generally refers to a device that comprises a lightproof chamber with an aperture fitted with a lens and a shutter through which the image of an object is projected onto a surface for translation into electrical impulses.

Although specific embodiments of and examples are described herein for illustrative purposes, various equivalent modifications can be made without departing from the spirit and scope of the invention, as will be recognized by those skilled in the relevant art.

FlexVisionTM 3D enables ABB robots to precisely locate the grip points of a disoriented object within a 3D space.

The software platform includes unique technologies such for easy calibration, with quick and reliable integration.

FlexVisionTM leverages world class vision technology as the most reliable and repeated VGR software for ABB robots.

Vision Guided Robotics Provides Savings • Manage variation in part styles and location • Eliminate costly precision fixturing, mechanical part crowding and dunnage • Automate operations that previously required human interaction • Increase up–time and eliminate robot crashes by seeing the part on racks The FlexVisionTM VGR Platform • Multi–camera 3D (GigE / PoE) • Single camera 2D • Robot mounted and stationary mounted cameras • Automatic camera calibration • Automated accuracy validation • Extremely fast set–up and calibration processes • Industrial, extreme flex cable system • 3D position of parts in full 6º of freedom Main Applications • Material handling • Machine tending • Dispensing & sealing • Press automation • Powertrain assembly • Body–in–White Technical Data | FlexVision TM 3D Vision Systems Supported Robot Types: Robot controller

Any reproduction, disclo- sure to third parties or utilization of its contents – in whole or in parts – is forbidden without prior written consent of ABB.

Integrated ABB vision is low maintenance and reliable With over 200 systems installed and seven years of continuous design innovations, TrueView is a reliable and robust vision solution for ABB robots.

ABB provides standard VGR modules to save robot program- ming time, and a TrueView FlexPendent GUI to provide factory workers with a simplified access point for trouble- shooting and calibration.

Vision guided robotics provides savings Manage variation in part styles and part location Eliminate costly precision fixturing, mechanical part crowding and dunnage Automate operations that previously required human interaction Increases “Up-Time” and eliminates robot crashes by seeing the part on racks Enhances quality using basic inspection and part identification techniques TrueView 5.12™ Vision Guided Robotics Copyright, ROB0079EN_B August 2009 Produced in USA ABB reserves the right to change specifications without notice.

TrueView 5.12 Vision Platform Xi2DTM • Single or multi camera 2D information in 3 degrees of freedom (x, y, Rz) Error proofing using feature present / absence and part identification • Calculate tolerances and positioning data for quality checks •

+/- 1 mm (part dependent) .5 sec for 2D applications 1 – 3 seconds for 2.5D and 3D applications +/- 15 degrees, +/- 300 mm Sony CCD cameras (analog) JAI and Baumer GigE cameras (digital) 8 or 12mm (standard) Red LED light bars (8” or 12” standard) with analog control by robot 180W Power-Supply - DDR II 1G 667MHZ memory Intel Core 2 Duo T7500@2.2