Vehicle Scenario-Matched Data Return for Faster Autonomous Driving Updates
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data collection methods for autonomous driving vehicles result in slow data return, leading to poor algorithm optimization and insufficient accuracy in decision-making, due to limited scenario data and inflexible parameter settings.
Innovation Solution
A method involving real-time scenario matching and data upload based on cloud server-provided configuration items, including scenario and return parameters, allowing flexible and accurate data return without relying on OTA technology.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If data collection is implemented through irregular upgrade and iteration of software on the vehicle, then new scenario data collection can be achieved, but data return speed is slow and algorithm optimization is delayed
Solution Approach 1:
The patent implements dynamic scenario matching where the system continuously monitors vehicle operating conditions and dynamically matches them against predefined scenario parameters. This allows the data collection system to adapt in real-time to new scenarios without requiring software updates, thereby maintaining high data return speed while improving adaptability to new situations.
Solution Approach 2:
The system changes parameters by allowing flexible configuration of scenario parameters and return parameters through the cloud server. This enables the system to adjust data collection priorities, thresholds, and types based on specific scenario needs, improving both adaptability to new scenarios and the efficiency of data return without requiring irregular software iterations.
2Productivity
If data collection parameters are fixed when the vehicle leaves the factory, then initial data collection can be performed, but the system cannot adapt to new scenarios and algorithm optimization is hindered
Solution Approach 1:
The patent implements a feedback mechanism where the cloud server receives uploaded scenario data, analyzes it, and generates updated scenario parameters and return parameters that are sent back to the vehicle. This closed-loop feedback system allows the vehicle to continuously improve its data collection capabilities for new scenarios while maintaining efficient data collection operations, resolving the contradiction between fixed initial parameters and the need for scenario adaptability.
Solution Approach 2:
The system performs preliminary action by pre-configuring multiple scenario parameters and return parameters when the vehicle leaves the factory. These pre-configured parameters cover various possible scenarios, allowing the vehicle to immediately begin adaptive data collection for multiple scenario types without requiring immediate software updates, thus maintaining both initial productivity and scenario adaptability.
3Quantity of substance
If all scenario data is collected and uploaded continuously, then comprehensive data is obtained, but data validity is reduced and storage resources are wasted
Solution Approach 1:
The patent applies local quality by implementing scenario-specific data collection strategies where different scenario parameters dictate what data is collected, when it is collected, and how it is prioritized. Each scenario has customized return parameters that determine data collection frequency, thresholds, and types, ensuring that only relevant high-validity data is collected for each specific scenario rather than uniformly collecting all possible data, thus improving data validity while maintaining appropriate data volume.
4Productivity
If data collection frequency is increased to improve algorithm optimization, then more scenario data is obtained, but energy consumption increases and system complexity grows
Solution Approach 1:
The patent implements periodic action by setting configurable collection frequencies and time intervals for different scenario parameters. The system collects and uploads data at optimized intervals based on scenario importance, data change rates, and priority levels rather than continuously or at fixed high frequencies. This periodic data collection approach accelerates algorithm optimization by focusing on critical scenarios while reducing overall energy consumption compared to uniform high-frequency collection.
Data Source
Figure 1~2
Figure 3~5
AI summary
The disclosure relates to the field of autonomous driving technologies, and specifically provides a method and system for returning data on the vehicle, a controller on the vehicle, a cloud server, and a vehicle, the cloud server can send data configuration items according to actual requirements, the data configuration items including scenario parameters and return parameters related to the scenario parameters, obtain a real-time scenario of a vehicle ,match the scenario with the scenario parameters, and upload matched data on the vehicle to the cloud server. The data on the vehicle can be flexibly, quickly, and accurately uploaded to the cloud server according to the requirements of the cloud server and can be returned without using the over-the-air (OTA) technology, which can improve the data validity more effectively and accelerate the optimization of algorithms for autonomous driving.