Amazon.com, Inc. v. SpeakWare, Inc.

1. Case Identification

• Case #: IPR2019-00999
• Patent #: 6,397,186
• Filed: April 19, 2019
• Petitioner(s): Amazon.com, Inc.
• Patent Owner(s): Speakware, Inc.
• Challenged Claims: 1-39, 41-55

2. Patent Overview

• Title: Audio Signal Activated Control System
• Brief Description: The ’186 patent discloses systems and methods for a voice-operated remote control for appliances. The core inventive concept asserted is the ability to switch from a low-power "sleep mode" to an "active" speech recognition mode based on the amplitude of a detected audio signal, avoiding the need for a manual "talk switch."

3. Grounds for Unpatentability

Ground 1: Obviousness over Lett/Houser - Claims 1-4, 7-16, 19-21, 23-26, 28-36, 39, 41, 43-52, and 55 are obvious over Lett/Houser.

• Prior Art Relied Upon: Houser (Patent 5,774,859) and Lett (Patent 5,771,064). Petitioner argued these references constitute a single prior art document ("Lett/Houser") because Lett explicitly incorporates the application underlying Houser by reference.
• Core Argument for this Ground:
  • Prior Art Mapping: Petitioner argued Houser taught all key elements of the independent claims. Houser described a speech interface for controlling appliances (e.g., VCRs) that used a "threshold element" to "sense when the sound level exceeds a certain level" and enable speech recognition components. This directly maps to the ’186 patent's central feature of switching from a low-power sound activation mode to a speech recognition mode based on signal amplitude. Lett provided a modular system context, describing an "applications module" with its own processor for features like a speech interface, which would be implemented as described in Houser.
  • Motivation to Combine: The primary motivation to place the processor into a low-power state, as required by the claims, arose from Houser's express teaching of power-saving. While Houser disclosed deactivating "components," Petitioner argued it would have been obvious to a person of ordinary skill in the art (POSITA) to include the power-intensive main processor among those components to maximize power savings. Processors of the era, such as the PowerPC 603 mentioned in the art, were known to have "Sleep modes." Implementing Houser's interface in Lett's separate module would also make it easier to power down the speech processor without affecting other terminal functions.
  • Expectation of Success: A POSITA would have had a high expectation of success because implementing power-saving modes in processors was a well-known technique, and Houser already provided the fundamental architecture for amplitude-based activation.

Ground 2: Obviousness over Lett/Houser and Stanley - Claims 5-6, 22, 27, and 42 are obvious over Lett/Houser in view of Stanley.

• Prior Art Relied Upon: Lett/Houser in combination with Stanley (Patent 5,684,924).
• Core Argument for this Ground:
  • Prior Art Mapping: The challenged dependent claims added limitations requiring user-adjustable settings for the activation threshold (claim 5) and signal amplification (claim 6). Stanley, titled "User Adaptable Speech Recognition System," directly addressed this by teaching utility modules for adjusting "sensitivity" (the discrimination level/threshold) and "gain" (amplification). Stanley explicitly taught a user-adjustable pre-amplifier to vary the gain and a user-adjustable threshold to distinguish speech from background noise.
  • Motivation to Combine: A POSITA would combine Stanley with the Lett/Houser system to improve its performance and user-friendliness. Stanley explained that improperly tuned parameters adversely affect system accuracy and that providing adjustment modules is a typical solution. Adding Stanley's user-adjustable controls to the Lett/Houser system would allow users to optimize its performance for different noise environments, a clear and predictable improvement.
  • Expectation of Success: Success would be expected because Stanley provided a known solution to a known problem in speech recognition systems. The similarity in system architecture (microphone, front-end conditioning, processor) between Stanley and Lett/Houser ensured the combination would be straightforward.

Ground 3: Obviousness over Lett/Houser and Fischer - Claims 17-18, 37-38, and 53-54 are obvious over Lett/Houser in view of Fischer.

• Prior Art Relied Upon: Lett/Houser in combination with Fischer (WO 94/03017).
• Core Argument for this Ground:
  • Prior Art Mapping: These dependent claims added a "programming mode" to enable a user to define control signals, including associating a single audible command with a plurality of control signals (i.e., a macro). Fischer disclosed a "Universal remote control device" with a "SETUP" mode (the claimed "programming mode") for learning IR codes. Critically, Fischer taught using this mode to create macros where a single command (e.g., "PLAY") could automatically output a sequence of control codes to multiple devices (e.g., turn on TV, then send play command to VCR), directly teaching the "plurality of control signals" limitation.
  • Motivation to Combine: Houser acknowledged that a user might need to issue multiple separate commands to perform a single function, such as recording a program. Fischer was explicitly designed to solve this inefficiency by creating macros. A POSITA would have been motivated to incorporate Fischer's macro functionality into the Lett/Houser system to improve its usability and efficiency, a direct and recognized benefit.
  • Expectation of Success: The combination was predictable and expected to succeed. Both systems used voice commands, microprocessors, and IR transmitters to control appliances. Integrating Fischer's software-based macro functionality into Houser's terminal, where the necessary processor and IR components were already located, would have been a logical design choice.

4. Arguments Regarding Discretionary Denial

• Petitioner argued that discretionary denial under §314(a) or §325(d) would be inappropriate. It contended that the General Plastic factors strongly favored institution because Amazon had not previously filed any petition against the ’186 patent. Petitioner asserted that although it was a co-defendant with other petitioners (Google, Microsoft, Apple) in parallel litigation, co-defendants are distinct parties with distinct interests. Furthermore, Petitioner argued the prior art and grounds presented were new, not having been raised in other petitions or during original prosecution. Finally, it asserted the petition was simple and would not impose an excessive burden on the Board's resources.

5. Relief Requested

• Petitioner requested the institution of an inter partes review and the cancellation of claims 1-39 and 41-55 of Patent 6,397,186 as unpatentable.