Confidence Zone Method for Road Geometry Estimation Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for estimating road geometry in vehicles lack the ability to determine the confidence level of the accuracy of the estimated geometry, which is crucial for autonomous driving and safety applications, as they do not provide reliable information on the error magnitude of the estimates.
Innovation Solution
A method in a vehicle to determine the confidence level of an estimated road geometry by obtaining a reference position of a road object, updating a confidence zone with known reference confidence levels, and associating the confidence level based on whether the reference position is within the confidence zone, using on-board sensors and wireless communications to gather data from various sources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If road geometry is estimated using on-board sensors and wireless communications, then the vehicle can obtain position and geometry data for autonomous driving applications, but the reliability of the estimated geometry is insufficient because the error magnitude cannot be determined
Solution Approach 1:
The patent introduces confidence zones as an intermediary mechanism between the estimated road geometry and the autonomous driving control system. These confidence zones represent spatial regions where the estimated geometry meets specific accuracy criteria, acting as a mediator that translates uncertain geometric estimates into reliable decision-making information for safety-critical applications
Solution Approach 2:
The patent transforms the estimated road geometry by adding a new parameter dimension - the confidence level or accuracy metric. Instead of仅提供 geometric data, the system now provides geometry estimates accompanied by confidence indicators that quantify the reliability of each estimated parameter, enabling the control system to make informed decisions based on both position and uncertainty
2Reliability
If the confidence level determination method is made comprehensive and accurate, then the reliability of autonomous driving decisions is improved, but the computational complexity increases
Solution Approach 1:
The patent segments the road geometry estimation problem into multiple confidence zones with different accuracy characteristics. Instead of computing a single confidence level for the entire road geometry, the system divides the estimation space into zones where each zone has its own confidence metrics, allowing for more manageable and efficient computation while maintaining comprehensive reliability assessment
Solution Approach 2:
The patent implements confidence level determination selectively rather than uniformly across all road geometry estimates. The system focuses computational resources on determining confidence levels for critical sections of the road geometry that most impact autonomous driving safety, while using simplified or pre-computed confidence metrics for less critical areas, thereby reducing overall computational complexity
Data Source
AI summary
The present disclosure relates to estimated road geometries used in vehicles, and in particular to a method and a control unit in a vehicle for determining a confidence level of an estimated road geometry. A method for determining a confidence level of an estimated road geometry is disclosed, the road geometry being estimated at least partly based on a position of a road object. The method comprises obtaining a reference position of the road object relative to the vehicle, updating a reach of a confidence zone of the vehicle, wherein a part of the estimated road geometry comprised within the confidence zone is associated with one or more reference confidence levels, and determining the confidence level of the estimated road geometry based on whether the reference position is comprised in the confidence zone.