Shenzhen Zhiyi Technology Co Ltd v. iRobot Corp

1. Case Identification

• Case #: IPR2017-02078
• Patent #: 7,155,308
• Filed: September 8, 2017
• Petitioner(s): Shenzhen Zhiyi Technology Co. Ltd., d/b/a iLife
• Patent Owner(s): iRobot Corp.
• Challenged Claims: 1-3, 7, 11-12, 17, 19-20, 28, and 34

2. Patent Overview

• Title: Obstacle Detection System for Mobile Robot
• Brief Description: The ’308 patent discloses a sensor subsystem for an autonomous mobile robot. The system uses optical emitters and detectors, such as infrared (IR) sensors, to detect obstacles, particularly drop-offs like stairs or ledges (cliffs), enabling the robot to avoid falling.

3. Grounds for Unpatentability

Ground 1: Obviousness over Drunk-106 and Jones-MR - Claims 1-2, 7, 11-12, 19-20, and 28 are obvious over Drunk-106 in view of Jones-MR.

• Prior Art Relied Upon: Drunk-106 (Patent 5,377,106) and Jones-MR (a 1998 book, "Mobile Robots: Inspiration to Implementation").
• Core Argument for this Ground:
  • Prior Art Mapping: Petitioner argued that Drunk-106 discloses a basic autonomous cleaning robot with multiple sensors, including downward-facing IR sensors for cliff detection. However, these sensors are simple. Jones-MR, a robotics implementation guide co-authored by the ’308 patent's inventor, was argued to disclose a more sophisticated IR sensor (the GP2D02) that uses triangulation to determine range. This sensor was described as being less sensitive to surface color, texture, and ambient light. Petitioner asserted that substituting the superior GP2D02 sensor from Jones-MR for the basic IR sensors of Drunk-106 would arrive at the invention of the independent claims, including an optical emitter/detector pair with intersecting fields that provides an output to redirect the robot when a cliff is detected.
  • Motivation to Combine: A Person of Ordinary Skill in the Art (POSITA) would combine the teachings to improve the performance and reliability of the Drunk-106 robot. Replacing a simple, known sensor with a more sophisticated, commercially available, and well-documented sensor (GP2D02) to achieve more accurate distance measurements and navigational capabilities was presented as a simple and predictable design choice.
  • Expectation of Success: Petitioner contended a POSITA would have a high expectation of success because the substitution involved replacing one type of IR proximity sensor with another, more advanced one for the same purpose (cliff detection). The interface of the GP2D02 sensor was described as comparable to standard IR sensors, making integration straightforward and the operational outcome entirely predictable.

Ground 2: Obviousness over Drunk-106, Jones-MR, and Lee-291 - Claims 3, 17, and 34 are obvious over the combination of Drunk-106, Jones-MR, and Lee-291.

• Prior Art Relied Upon: Drunk-106 (Patent 5,377,106), Jones-MR (a 1998 book, "Mobile Robots: Inspiration to Implementation"), and Lee-291 (Patent 5,621,291).
• Core Argument for this Ground:
  • Prior Art Mapping: This ground built upon the combination in Ground 1 by adding Lee-291. Petitioner argued that challenged dependent claims 3, 17, and 34 add limitations related to wall-following behavior. While Drunk-106 required a precise, pre-programmed map to clean along walls, Lee-291 was asserted to explicitly teach a robotic vacuum with a sensor-based wall-following capability. Claim 3 requires redirecting the robot toward a wall when it is not detected, which Petitioner mapped to the wall-following behavior in Lee-291. Claim 17 requires operating in an obstacle-following mode for a specific distance (2x to 10x the robot's work width), which Petitioner argued was disclosed or made obvious by Lee-291’s disclosure of following a wall to a "target point."
  • Motivation to Combine: A POSITA would have been motivated to add the wall-following feature from Lee-291 to the improved Drunk-106/Jones-MR robot from Ground 1. This would enhance its functionality, ensuring it could thoroughly clean edges and corners of a room, a well-known desirable feature for cleaning robots, without requiring expensive pre-mapping.
  • Expectation of Success: Adding a known, well-understood behavior like wall-following to an autonomous robot was argued to be a routine task for a POSITA. The combination involved integrating known sensor types (from Jones-MR) with known control logic (from Lee-291) into a known robot platform (Drunk-106), which would have produced predictable results.

Ground 3: Obviousness over Drunk-106 and Everett - Claims 1-2, 7, 11-12, 19-20, and 28 are obvious over Drunk-106 in view of Everett.

• Prior Art Relied Upon: Drunk-106 (Patent 5,377,106) and Everett (a 1995 book, "Sensors for Mobile Robots: Theory and Application").
• Core Argument for this Ground:
  • Prior Art Mapping: This ground presented an alternative to Ground 1, substituting the Jones-MR reference with Everett. Petitioner argued that Everett, a reference book on mobile robot sensors, disclosed "convergent mode" diffuse proximity sensors. These sensors use a specific geometry where the emitter and detector are angled toward each other, creating a narrowly defined detection zone. This design was explicitly taught for detecting floor discontinuities (like stairs) and was described as being robust against variations in surface reflectivity, a known problem for simpler IR sensors. Petitioner asserted that replacing the basic IR sensors of Drunk-106 with the superior convergent-mode sensors from Everett would render the claims obvious.
  • Motivation to Combine: The motivation was identical to that in Ground 1: to improve the performance of the Drunk-106 robot's floor detection system. A POSITA would have recognized the known advantages of Everett's convergent-mode sensors and would have been motivated to incorporate them to create a more reliable and accurate cleaning robot.
  • Expectation of Success: Petitioner argued that success would have been expected because it was a straightforward substitution of one known sensor type for another to perform the same function. Since Drunk-106 already used IR sensors for floor detection, a POSITA would have possessed the basic engineering skills to implement the more accurate convergent-mode sensors taught by Everett, leading to a predictable improvement in performance.
• Additional Grounds: Petitioner asserted an additional challenge (Ground 4) that claims 3, 17, and 34 are obvious over Drunk-106 in view of Everett and Lee-291. This ground relied on the same rationale as Ground 2, but used the convergent-mode sensors from Everett (as detailed in Ground 3) instead of the GP2D02 sensor from Jones-MR as the basis for the wall-following sensor.

4. Relief Requested

• Petitioner requested that an inter partes review (IPR) be instituted and that claims 1-3, 7, 11-12, 17, 19-20, 28, and 34 of the ’308 patent be cancelled as unpatentable.