Automated Vehicle Localization Using Landmark Validation and Re-Localization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for localizing highly automated vehicles rely on GNSS systems, which are prone to fluctuations in accuracy due to quality issues, necessitating improved methods for precise vehicle positioning and surroundings awareness.
Innovation Solution
A method involving a localization module that performs global pose estimation using a combination of sensor measurements and digital map data, including landmark positions and properties, with error checking and potential re-localization through a particle-based approach when initial pose estimation is invalidated.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If GNSS-based localization is used for vehicle positioning, then global coverage and basic positioning capability are achieved, but localization accuracy fluctuates due to quality issues and environmental factors
Solution Approach 1:
The patent combines multiple localization sources (GNSS, map data, sensor measurements) into a unified localization system. The method integrates global pose estimation from GNSS with landmark detection from sensors and map matching, creating a hybrid localization approach that compensates for the weaknesses of individual systems and maintains accurate positioning even when GNSS quality degrades
Solution Approach 2:
The patent implements a feedback mechanism where the system continuously monitors the quality and consistency of localization data from multiple sources. When discrepancies or errors are detected in GNSS positioning, the system adjusts by relying more heavily on alternative sources such as map-matching and landmark-based localization, thereby maintaining reliable positioning accuracy
2Measurement precision
If detailed sensor measurements and landmark detection are performed to improve localization accuracy, then positioning precision increases, but system complexity and computational requirements increase
Solution Approach 1:
The patent pre-processes and stores landmark information from digital maps before runtime localization. Landmark positions and properties are extracted and organized in advance, allowing the system to quickly match sensor measurements against pre-prepared map data without performing complex real-time processing, thus reducing computational complexity while maintaining high positioning precision
Solution Approach 2:
The localization system is divided into independent modular components: GNSS processing module, landmark detection module, map-matching module, and fusion module. Each module operates independently and processes specific aspects of localization, making the overall system more manageable and less complex while achieving high precision through the coordinated work of specialized subsystems
Data Source
AI summary
A method for localizing a more highly automated vehicle (HAF), in particular a highly automated vehicle, in a digital map. The method includes: ascertaining a global pose estimation for the HAF using a localization module of a vehicle system of the HAF, the global pose estimation comprising a position and orientation of the HAF; transmitting at least one landmark position and at least one associated landmark property to the vehicle system; ascertaining a relative position of the landmark position concerning the HAF at least partially on the basis of the pose estimation and the landmark position; performing at least one sensor measurement and checking that the at least one landmark property is detectable at the relative position; and outputting an error indicator as the result of the checking. A corresponding system and a computer program are also described.