In-Vehicle AI Agent Handover During Network Disconnection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing human-agent interaction systems in vehicles face challenges in maintaining seamless communication with AI agents when network connectivity is disrupted, particularly in areas with poor communication environments.
Innovation Solution
Implementing a first AI agent on a vehicle's computer to interface with users locally and record conversation logs, enabling continued interaction processing when disconnected from the cloud-based second AI agent by switching to local interaction based on these logs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Power
If voice interaction agent inference processing is implemented on the cloud side to reduce calculation amount on vehicle, then processing power is improved, but system reliability deteriorates due to network disconnection risk
Solution Approach 1:
The system segments the voice interaction agent into two parts: a cloud-based inference processing component and a local edge component. The edge component stores conversation logs and can execute interaction processing locally when network disconnection occurs, while the cloud component handles inference processing when connected, thereby distributing functionality to balance processing power and reliability.
Solution Approach 2:
The system performs preliminary action by storing conversation logs locally on the edge device before network disconnection occurs. This pre-stored data enables the local AI agent to continue interaction processing without cloud connection, ensuring system reliability during network outages while maintaining cloud-based inference processing capability when available.
2Device complexity
If cloud-based AI agent is used to reduce in-vehicle computation resources, then device complexity is reduced, but continuity of interaction deteriorates during network disconnection
Solution Approach 1:
The system divides the AI agent functionality into cloud-based inference processing and edge-based interaction processing. The edge device maintains a local AI agent capable of executing interaction processing using stored conversation logs, ensuring continuous interaction during network disconnection while keeping in-vehicle device complexity relatively low.
Solution Approach 2:
The local AI agent on the edge device acts as an intermediary between the user and the cloud-based inference processing system. When the network is disconnected, this intermediary continues to handle interaction processing locally using stored conversation logs, thereby maintaining continuity of interaction without requiring direct cloud connection.
3Reliability
If conversation logs are recorded locally on the first computer, then reliability during disconnection is improved, but data loss risk increases if the first computer fails
Solution Approach 1:
The system applies local quality by storing conversation logs specifically on the edge device (first computer) where they are needed for local interaction processing during network disconnection. This localized storage ensures reliability during disconnection while the system architecture accepts the inherent data loss risk as a trade-off for enabling offline functionality.
Data Source
AI summary
An information processing method is executed by a first computer installed in a vehicle in a human-agent interaction system. The method includes interfacing interaction processing with a user by a first AI agent implemented on a second computer on a network by using a microphone and a speaker provided in the vehicle. The interfacing is performed while being connected to the second computer via a communication device of the vehicle. The method includes recording a first conversation log related to conversation content between the first AI agent and the user in a memory of the first computer. The method includes, when connection to the second computer is disconnected, causing the communication terminal or a second AI agent implemented in the first computer to execute second interaction processing with the user based on the first conversation log.