Server-Selected Recognition Model for Vehicle Scene Data Collection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data collection systems for connected cars cannot collect data when the vehicle is in specific scenes, as they rely on sensor values exceeding threshold conditions, limiting data collection efficiency.
Innovation Solution
An information collection system where a server selects and transmits recognition models to vehicles via a network, enabling the vehicle to determine if it is in a specific scene and transmit relevant data to the server, allowing targeted data collection based on sensor information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If data collection is triggered by sensor values exceeding threshold conditions, then data collection efficiency is improved, but data collection cannot occur when the vehicle is in specific scenes
Solution Approach 1:
The server performs preliminary actions by selecting and transmitting recognition models to vehicles before data collection is needed. The recognition models are prepared in advance and sent to vehicles, enabling them to autonomously determine whether they are in specific scenes and trigger data collection accordingly, thus resolving the contradiction between efficiency and coverage
Solution Approach 2:
Recognition models serve as intermediaries between the server and vehicles. The server transmits these models to vehicles, which then use them to autonomously determine scene presence and trigger data collection. This intermediary mechanism enables the server to indirectly control data collection based on specific scenes without continuous monitoring, improving both efficiency and coverage
2Reliability
If all data is collected from in-vehicle apparatus, then complete data coverage is achieved, but data collection efficiency decreases
Solution Approach 1:
The system extracts only the necessary data collection triggers (specific scene recognition) from the comprehensive data collection approach. By using recognition models to identify specific scenes, the system extracts only the relevant data collection opportunities, avoiding the inefficiency of collecting all data while ensuring complete coverage of important scenarios
Solution Approach 2:
The system changes the parameter for triggering data collection from continuous monitoring of all sensor values to discrete scene recognition based on transmitted models. This parameter change transforms the data collection approach from exhaustive to targeted, improving efficiency while maintaining reliability through accurate scene identification
Data Source
AI summary
A recognition model selection means selects a recognition model for identifying that a vehicle is in a situation corresponding to a specific scene on the basis of sensor information. A transmission means transmits the selected recognition model to a vehicle. A scene determination means determines whether or not the vehicle is in a situation corresponding to a specific scene on the basis of the recognition model received from a server and sensor information. A data transmission means transmits information to the server in a case where the vehicle is determined to be in a situation corresponding to a specific scene. A data collection means collects information transmitted from the vehicle.