V2X Roadway Feature Extraction for Faster High-Accuracy Maps
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing solutions for generating high-definition maps for automated vehicles are inadequate due to the large amount of data required, which leads to slow data upload and analysis times, making them insufficiently dynamic and unable to accurately reflect changing roadway features.
Innovation Solution
A machine learning system and learning client that extract roadway features from Vehicle-to-Everything (V2X) data using vehicle clustering analysis, enabling continuous updates and improved accuracy of maps by aggregating and analyzing smaller V2X data sets, forming a feedback loop to enhance feature extraction over time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If large amounts of data are used for map generation, then map accuracy is improved, but data upload and analysis time increase
Solution Approach 1:
The patent extracts only the essential roadway feature information from V2X data through vehicle clustering analysis, rather than processing all raw data. This extraction approach maintains map accuracy by identifying key patterns in vehicle trajectories while significantly reducing the data volume that needs to be uploaded and processed centrally.
Solution Approach 2:
The patent segments the map generation process into distributed vehicle clustering analysis performed locally by vehicles and a centralized aggregation step. This segmentation allows parallel processing of data at the vehicle level, reducing the time burden on the central system while maintaining comprehensive map accuracy.
2Quantity of substance
If prior map data is not updated, then data storage is reduced, but map accuracy deteriorates
Solution Approach 1:
The patent implements a feedback mechanism where vehicles continuously transmit V2X data and receive updated map data in return. This feedback loop enables the system to maintain current map accuracy by incorporating real-time roadway feature changes while managing data storage through selective updating based on detected changes.
3Adaptability or versatility
If traditional map generation methods are used, then comprehensive coverage is achieved, but system complexity increases
Solution Approach 1:
The patent enables vehicles to perform self-service vehicle clustering analysis on their own V2X data to extract roadway features locally. This self-service approach distributes the processing complexity to individual vehicles rather than requiring a complex centralized processing system, while still achieving comprehensive map coverage through aggregation of results from multiple vehicles.
Data Source
AI summary
The disclosure includes embodiments for extracting roadway features from Vehicle-to-Everything (V2X) data to build a map having an improved accuracy. In some embodiments, a method for a connected vehicle includes transmitting, via a communication unit of the connected vehicle, a V2X message that includes V2X data. The method includes receiving, via the communication unit, map data describing a map that depicts one or more roadway features in a roadway region of the connected vehicle. The one or more roadway features are generated through a vehicle clustering analysis based on an aggregation of the V2X data in the roadway region so that an accuracy of the map is improved. The method includes modifying an operation of a vehicle control system of the connected vehicle based on the map data to improve the operation of the connected vehicle by reducing a risk of a collision involving the connected vehicle.