Vehicle Localization Using Map Feature Change Probabilities
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current vehicle localization systems rely on digital maps that may become outdated, leading to inaccurate vehicle positioning, especially in autonomous driving modes where timely and accurate location information is crucial.
Innovation Solution
The method enhances vehicle localization by assigning attributes such as cycle duration and time of last change to features in the digital map, allowing for probability calculations of changes, which helps identify and address outdated data, thereby increasing localization robustness and accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If digital map data is used for vehicle localization, then localization can be achieved by comparing sensor-detected features with map features, but the map data may become outdated leading to inaccurate vehicle positioning
Solution Approach 1:
The system performs preliminary actions by continuously monitoring feature change probabilities and detecting outdated map data before it significantly degrades localization accuracy. The probabilistic model predicts when map features may become obsolete, allowing the system to proactively identify and flag outdated data rather than waiting for localization errors to manifest.
Solution Approach 2:
The system implements feedback mechanisms by continuously comparing sensor-detected features with map features, calculating change probabilities, and using this information to assess map data currency. This closed-loop feedback allows the system to adapt to changing environmental conditions and maintain localization accuracy despite map data aging.
2Measurement precision
If more features are included in localization to improve precision, then localization accuracy increases, but the probability that features are no longer entirely up-to-date increases
Solution Approach 1:
The system applies local quality by assigning different change probability characteristics to different map features based on their specific properties and environmental contexts. Rather than treating all features uniformly, the system evaluates each feature's likelihood of change individually, allowing high-precision localization using multiple features while accounting for their varying currency through probabilistic assessment.
3Productivity
If digital map data is used for localization, then vehicle positioning can be achieved, but the system cannot quickly detect faulty or outdated maps
Solution Approach 1:
The system performs self-service by autonomously assessing the quality and currency of its own map data through probabilistic analysis of feature changes. The localization system itself generates the information needed to detect outdated maps, eliminating the need for external validation systems and enabling rapid self-diagnosis of map data quality issues.
Data Source
Figure 1
Figure 2~3
AI summary
The invention relates to a method and an associated system for vehicle localisation using a digital map (12). According to the method, the digital map (12) allocates at least one pre-determined attribute (13) to the features (4), said attributes characterising any actual changes (A) to the features (4) that may occur in comparison with the currently present digital map (12). Furthermore, respective probabilities (P) for the changes (A) are determined on the basis of the allocated attributes (13) and the vehicle is localised taking into account the determined probabilities (P).