Roku Inc v. Universal Electronics Inc

1. Case Identification

• Case #: IPR2022-00943
• Patent #: 10,930,276
• Filed: April 29, 2022
• Petitioner(s): Roku, Inc.
• Patent Owner(s): Universal Electronics Inc.
• Challenged Claims: 1-16

2. Patent Overview

• Title: Adjusting Audio Output by a Controllable Appliance
• Brief Description: The ’276 patent discloses a method for a voice-controlled device to automatically adjust the volume of a connected appliance, such as a television. The method involves establishing a noise threshold based on environmental sound and, upon receiving a user's speech input, lowering the appliance's volume if the current environmental noise level exceeds that threshold to improve voice command recognition.

3. Grounds for Unpatentability

Ground 1: Obviousness over Hart-787, FU, and Rosenberg - Claims 1-11 and 13-16 are obvious over Hart-787 in view of FU and further in view of Rosenberg.

• Prior Art Relied Upon: Hart-787 (Patent 9,251,787), FU (Application # 2017/0126192), and Rosenberg (Patent 9,509,269).
• Core Argument for this Ground:
  • Prior Art Mapping: Petitioner argued that the primary reference, Hart-787, discloses a voice-controlled device (e.g., an Amazon Echo) that can reduce the volume of a wirelessly-linked audio-producing appliance to improve the accuracy of its speech recognition. However, Hart-787 broadly teaches this capability but leaves the specific implementation of when and how to attenuate the audio to a person of ordinary skill in the art (POSITA). Petitioner asserted that FU supplies the missing element by teaching a voice control device that adjusts the volume of a connected device if the detected environmental noise level is greater than a predetermined threshold. To improve upon FU's static threshold, Petitioner pointed to Rosenberg, which teaches establishing a dynamic noise threshold based on the current ambient environmental noise to prevent false triggers. The combination of these references allegedly teaches all limitations of independent claim 1, including using a first sound capture to establish a noise threshold (Rosenberg), receiving a speech input (Hart-787), using a second sound capture at the time of the input to determine a current noise level (Hart-787 as modified by FU), comparing the level to the threshold (FU), and issuing a command to lower the appliance volume if the threshold is exceeded (Hart-787).
  • Motivation to Combine: A POSITA would combine FU's threshold-based volume adjustment with Hart-787's system to achieve Hart-787's stated goal of improving speech recognition more effectively. This would prevent the system from unnecessarily lowering the volume when ambient noise is low. Furthermore, a POSITA would incorporate Rosenberg’s dynamic threshold, as it was a known and advantageous alternative to a static threshold, making the system more robust in environments with variable noise levels, which is a predictable improvement.
  • Expectation of Success: Petitioner contended a POSITA would have a reasonable expectation of success because the combination involves applying known, simple, and compatible techniques (threshold comparisons and dynamic thresholds) to the known system of Hart-787 to achieve the predictable result of improved voice command recognition in noisy environments.

Ground 2: Obviousness over Hart-787, FU, Rosenberg, and Hart-286 - Claim 12 is obvious over Hart-787 in view of FU and further in view of Rosenberg and Hart-286.

• Prior Art Relied Upon: Hart-787 (Patent 9,251,787), FU (Application # 2017/0126192), Rosenberg (Patent 9,509,269), and Hart-286 (Patent 9,466,286).
• Core Argument for this Ground:
  • Prior Art Mapping: This ground addresses dependent claim 12, which adds the limitation of associating the established noise threshold with a specific "time of day" and only performing the noise level comparison when a speech input is received during that associated time. Petitioner argued that the base combination of Hart-787, FU, and Rosenberg teaches the underlying method, and Hart-286 provides the missing time-of-day feature. Hart-286 discloses a voice-controlled device that uses different speech recognition thresholds at different times of day, based on usage patterns or expected background noise, to improve accuracy.
  • Motivation to Combine: A POSITA seeking to further refine the system of Ground 1 would be motivated to incorporate the teachings of Hart-286. Applying different noise thresholds based on the time of day was a known technique for improving the accuracy and reliability of speech recognition systems by adapting to predictable fluctuations in environmental conditions (e.g., a louder household in the evening vs. a quieter one in the morning). This modification would yield the predictable benefit of a more accurate and context-aware system.
  • Expectation of Success: The integration of time-of-day-based thresholds from Hart-286 into the base combination was presented as a straightforward application of a known technique to enhance a system's performance in a predictable manner.

6. Arguments Regarding Discretionary Denial

• Petitioner argued that discretionary denial under 35 U.S.C. §325(d) is inappropriate because the primary references asserted in the petition (Hart-787, FU, Rosenberg, and Hart-286) were not considered by the examiner during the original prosecution of the ’276 patent.
• Petitioner further argued that discretionary denial under §314(a) based on Fintiv factors is inapplicable because, while the parties are involved in other litigation, the Patent Owner has not asserted the ’276 patent against Petitioner in any co-pending district court case.

7. Relief Requested

• Petitioner requests institution of an inter partes review and cancellation of claims 1-16 of Patent 10,930,276 as unpatentable.