Voice Agent Utterance Timing Control for Overlap Prevention
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In environments with multiple voice agents, overlapping utterances and interruptions between agents and users can impair the user experience, as existing solutions struggle to prioritize important messages and interactions effectively, especially when tasks and message importance are diverse and dynamic.
Innovation Solution
An information processing apparatus and method that includes a sending unit, receiving unit, and control unit to manage utterance timing based on interactive information exchanged between voice agents, allowing voice agents to stand by or postpone utterances when another agent is interacting with the user, and prioritizing the agent that started interacting first to avoid overlaps and interruptions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If multiple voice agents operate independently in the same space, then each agent can provide voice services autonomously, but utterances overlap and user experience deteriorates
Solution Approach 1:
The patent introduces an utterance timing control unit as an intermediary mechanism that mediates between multiple voice agents. This controller receives interactive information from other voice agents and adjusts the utterance timing of the own voice agent accordingly, preventing overlapping utterances while maintaining autonomous operation. The mediator coordinates timing without requiring centralized control of all agents.
Solution Approach 2:
The voice agent performs preliminary actions by sending interactive information about its utterance plans before actually uttering. Other agents receive this information in advance and adjust their timing accordingly. This preliminary exchange of information allows agents to coordinate without direct real-time interference, maintaining autonomy while avoiding overlaps.
2Reliability
If voice agents prioritize important messages, then critical information can be delivered, but determining importance becomes complex when tasks are diverse and dynamic
Solution Approach 1:
The patent changes the parameter for determining utterance priority from complex semantic analysis of message importance to a simpler temporal parameter: which agent started interacting first. This parameter change simplifies the determination mechanism while still ensuring reliable delivery of critical information, as the first-interacting agent naturally receives priority without requiring complex importance evaluation.
3Object-affected harmful factors
If a voice agent waits for interactive information from other agents, then utterance overlaps are prevented, but response time increases
Solution Approach 1:
The voice agent performs partial waiting by sending interactive information only when necessary to prevent overlaps, rather than continuously waiting for all other agents. The agent waits minimally for the duration needed to exchange timing information with other agents, then proceeds with its own utterance plan. This partial action approach prevents overlaps while minimizing time loss.
Data Source
AI summary
Provided is an information processing apparatus that has an utterance function or controls the utterance function. The information processing apparatus includes a sending unit that sends interactive information regarding a voice agent, a receiving unit that receives interactive information regarding another voice, and a control unit that controls an utterance timing of the voice agent on the basis of the interactive information regarding another voice agent received by the receiving unit. The control unit causes utterance by the voice agent to stand by on the basis of the interactive information received from another voice agent. Moreover, the control unit causes the interactive information to be continuously sent during the utterance by the voice agent and during interaction between the voice agent and a user.