Vehicle AI Data Saving for In-Vehicle Memory Limits
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The limited capacity of in-vehicle memory devices poses a challenge when saving data related to AI-driven autonomous driving control, as the amount of data involved is substantial, and there is a need for an efficient method to determine which data to save and how to manage it effectively.
Innovation Solution
A method and device that determine whether AI is involved in computer processing for autonomous driving control, and if a predetermined save condition is met, save input and output data within a specific time frame in the vehicle's memory device, thereby optimizing data storage and preventing memory device capacity insufficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If all data related to AI-driven autonomous driving control is saved in the in-vehicle memory device, then complete data for evaluation and verification is obtained, but the memory device capacity is insufficient
Solution Approach 1:
The patent segments the autonomous driving control data into multiple categories: sensor data, recognition results, planning results, control results, and AI model information. This segmentation allows selective saving of only essential data types required for AI evaluation, preventing memory overflow while maintaining data completeness for verification purposes.
Solution Approach 2:
The patent extracts and saves only the essential input and output data required for AI evaluation, rather than saving all generated data. Specifically, it extracts sensor data, recognition results, planning results, control results, and AI model information, while excluding redundant intermediate processing data, thereby reducing storage requirements while maintaining evaluation capability.
2Reliability
If data is saved for long periods to ensure complete evaluation coverage, then comprehensive verification is achieved, but memory device free space becomes insufficient
Solution Approach 1:
The patent performs preliminary determination of essential data types and retention periods before actual data generation. It pre-establishes what data must be saved (sensor data, recognition results, planning results, control results, AI model information) and for how long, allowing the system to selectively save only this predetermined essential data, thereby ensuring evaluation coverage while controlling storage consumption.
3Quantity of substance
If selective data saving is implemented to preserve memory space, then free space is maintained, but data completeness for AI evaluation may be compromised
Solution Approach 1:
The patent creates a universal data saving framework that handles multiple data types (sensor data, recognition results, planning results, control results, AI model information) through a single standardized process. This multi-functional approach ensures that all essential data types required for AI evaluation are captured uniformly, preventing information loss while maintaining memory efficiency through consistent selective saving criteria.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A method for managing data related to a vehicle (1) in which autonomous driving control is performed is provided. In the method, it is determined whether artificial intelligence is involved in computer processing for performing the autonomous driving control (S14). In the method, it is also determined whether a predetermined save condition is satisfied (S15). When it is determined that the artificial intelligence is involved in the computer processing and it is determined that the predetermined save condition is satisfied, input and output data of the computer processing within a predetermined save time (Tsave) is saved based on a timing at which it is determined that the predetermined save condition is satisfied in a memory device (120) of the vehicle (S16).