Obviousness

Obviousness Analysis under 35 U.S.C. § 103

The Patent Trial and Appeal Board (PTAB) has already issued a Final Written Decision (FWD) in IPR2023-00119, finding claims 1-18 of US10593330 unpatentable under 35 U.S.C. §§ 102 and/or 103. [cite: https://portal.unifiedpatents.com/ptab/case/IPR2023-00119] This decision, issued on 2024-06-14, explicitly states that Unified Patents, LLC demonstrated "by a preponderance of the evidence that claims 1-18 of the ’330 patent are unpatentable." [cite: https://portal.unifiedpatents.com/ptab/case/IPR2023-00119] Google LLC has appealed this decision to the Court of Appeals for the Federal Circuit (CAFC) under case number 2025-1033. [cite: https://portal.unifiedpatents.com/ptab/case/IPR2023-00119]

Prior Art References Considered by the PTAB:

The PTAB instituted review of claims 1-18 based on obviousness grounds, considering various combinations of the following prior art references [cite: https://portal.unifiedpatents.com/ptab/case/IPR2023-00119]:

• US 2013/0246067 A1 (Kopani)
• US 2012/0066008 A1 (Sinha)
• US 2008/0249780 A1 (Kurosaki)
• US 2013/0066657 A1 (Bostick)
• US 2013/0185078 A1 (O'Neill)

Motivation to Combine Prior Art References:

While the detailed contents of each specific prior art reference (Kopani, Sinha, Kurosaki, Bostick, and O'Neill) are not provided in the prompt, the PTAB's finding of unpatentability under 35 U.S.C. § 103 implies that a person having ordinary skill in the art (PHOSITA) would have found sufficient motivation to combine these references to arrive at the claimed invention of US10593330.

Based on the problem US10593330 aims to solve—namely, coordinating hotword detection among multiple devices to ensure only one device responds—a PHOSITA would likely be motivated to combine prior art references that address different aspects of this technical challenge. Common motivations for such combinations in the field of speech recognition and multi-device interaction include:

1. Addressing Redundancy and Conflicting Responses: As stated in US10593330, "when a single user has multiple devices trained to respond to their voice... it may be desirable to suppress responding to hotwords on devices that are not likely to be the ones a user intends to address." A PHOSITA would be motivated to combine existing hotword detection systems with mechanisms for inter-device communication and arbitration to prevent multiple devices from simultaneously responding or performing redundant actions.
2. Improving Efficiency and User Experience in Multi-Device Environments: In a "speech-enabled environment" with a "network of connected microphone devices," as described in US10593330, efficient management of device responses is critical. Prior art disclosing basic hotword detection (e.g., Kopani or Sinha focusing on speech recognition or hotword spotting) could be combined with prior art related to networked computing devices (e.g., Kurosaki, Bostick, or O'Neill potentially teaching device identification, communication protocols, or task distribution in a network).
3. Optimizing Resource Usage: Having multiple devices perform full speech recognition when only one is intended to respond wastes computational resources and battery life. A PHOSITA would seek to combine a hotword detection technique with a mechanism to select a single "best" device to proceed with full speech recognition, thereby conserving resources on other devices. This would involve comparing metrics (like hotword confidence scores, as taught in US10593330) across devices and having non-selected devices enter an inactive state.
4. Leveraging Existing Technologies: Many of the individual components of the claimed invention (hotword detection, inter-device communication, score comparison, and conditional activation) were known in the art. A PHOSITA would be motivated to combine these known elements in a logical way to solve the recognized problem of multi-device hotword collision. For example, if one reference teaches robust hotword detection and another teaches a method for devices in a local network to share data and arbitrate tasks, a PHOSITA would find it obvious to combine these to create a coordinated hotword detection system.

The PTAB's Final Written Decision confirms that, after reviewing the arguments and evidence presented by Unified Patents, they found these motivations and the resulting combinations of the cited prior art sufficient to render claims 1-18 of US10593330 unpatentable as obvious to a PHOSITA. The specific combinations and detailed reasoning for each claim are elaborated in the full Final Written Decision for IPR2023-00119, which can be accessed via the USPTO PTAB E2E portal. [cite: https://portal.unifiedpatents.com/ptab/case/IPR2023-00119]