Centralized Speech Processing for Multi-Assistant Handoffs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current speech processing systems face challenges in managing and coordinating interactions between multiple virtual assistants, leading to undesirable user experiences due to technical issues during handoffs between command processing subsystems, such as incorrect handling of utterances and lack of privacy protection, especially when different assistants have different speech processing architectures and permissions.
Innovation Solution
A centralized speech-processing system is introduced, featuring a plan determination component that orchestrates handoffs and command execution, using post-NLU data to determine ranked lists of plans for executing utterances, and a handoff controller to fill in data for customized responses, ensuring seamless transitions and user experience customization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If multiple virtual assistants with different speech processing architectures are integrated into a single system, then the system's adaptability and versatility improve, but the device complexity and coordination difficulty increase
Solution Approach 1:
The patent introduces a centralized speech processing system that acts as an intermediary between multiple virtual assistants and the user. This central system receives utterances, determines which assistant should handle them, and coordinates handoffs between different command processing subsystems. The intermediary architecture allows multiple assistants with different architectures to coexist without direct complex interactions, resolving the coordination difficulty while maintaining adaptability.
Solution Approach 2:
The system segments the speech processing functionality into separate command processing subsystems, each dedicated to a specific virtual assistant. This segmentation allows each assistant to maintain its own speech processing architecture independently while being managed through a centralized coordination layer, thus preserving versatility without increasing overall system complexity.
2Productivity
If handoffs between command processing subsystems are implemented, then the system's ability to route utterances to appropriate assistants improves, but technical issues during handoffs and reliability worsen
Solution Approach 1:
The centralized speech processing system performs preliminary determination of the appropriate command processing subsystem before the actual handoff occurs. By pre-determining the target assistant based on utterance analysis, the system prepares the handoff in advance, ensuring smooth transitions and reducing technical issues during the actual transfer of control between subsystems.
Solution Approach 2:
The system implements feedback mechanisms where the centralized speech processing system monitors the handoff process between command processing subsystems. When technical issues are detected during handoffs, the feedback loop allows the system to adjust and correct the routing decisions, thereby improving handoff reliability while maintaining efficient utterance routing.
3Reliability
If a centralized speech processing system is implemented to coordinate multiple assistants, then feedback and coordination among components improve, but the system requires more sophisticated control mechanisms increasing complexity
Solution Approach 1:
The centralized speech processing system is designed as a universal control mechanism that handles multiple functions: receiving utterances, determining appropriate assistants, coordinating handoffs, and monitoring system performance. By consolidating these diverse control functions into a single multi-functional system, the patent improves feedback and coordination without proportionally increasing overall system complexity.
Data Source
AI summary
A multi-assistant speech-processing system that centrally determines multiple execution plans to respond to a user input. A central component determines whether a particular input should be processed using a requested assistant or a different assistant or should be terminated. Assistant handoff may be determined based on system policies as well as user input-specific data. A ranked list of execution options may be supplemented by augmented data corresponding to messages to a user. The system may attempt to execute plans in the ranked order until a plan succeeds.