HD Map Localization Variant Selection for Autonomous Driving Contexts
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles face challenges in accurate localization due to varying geographical contexts and sensor data quality, leading to inconsistent performance across different regions and conditions, which can result in navigation failures.
Innovation Solution
A system that uses a localization variant index to select the most suitable localization technique based on driving contexts, integrating sensor data from multiple modalities like camera images, lidar scans, and GPS using Kalman filtering to ensure accurate and efficient navigation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a single localization technique is used, then the system is simple to implement, but the localization accuracy deteriorates in varying geographical contexts
Solution Approach 1:
The patent segments the localization system into multiple independent localization techniques (GPS-based, IMU-based, lidar-based, camera-based), each optimized for specific driving contexts. The localization variant index divides the operational space into different contexts and selects appropriate techniques for each, resolving the contradiction by making the system adaptable without requiring all techniques to operate simultaneously in all conditions.
Solution Approach 2:
The patent implements dynamic selection of localization techniques through the localization variant index, which adapts the localization approach based on current driving context (geographical region, speed, time of day, weather). This dynamic adaptation allows the system to maintain high accuracy across varying conditions without the complexity of a static multi-technique system, as only context-relevant techniques are activated.
2Measurement precision
If multiple localization techniques are used, then the localization accuracy improves across different contexts, but the system complexity increases
Solution Approach 1:
The patent applies preliminary action by pre-evaluating and storing the performance characteristics of multiple localization techniques across different driving contexts in the localization variant index before actual operation. This offline preparation allows the runtime system to simply query and select the best technique for the current context, achieving high accuracy without the runtime complexity of evaluating multiple techniques simultaneously.
Solution Approach 2:
The localization variant index serves as an intermediary between the multiple localization techniques and the vehicle navigation system. It abstracts the complexity of multiple techniques by providing a unified interface that selects and manages the appropriate technique based on driving context, thereby improving localization accuracy while shielding the overall system from the complexity of handling multiple techniques.
3Reliability
If localization parameters are tuned for one geographical region, then the localization performance is optimized for that region, but the performance deteriorates in other regions
Solution Approach 1:
The patent applies local quality by storing region-specific performance characteristics and optimal parameter settings in the localization variant index for different geographical regions. Each region has its own tuned parameters and preferred localization techniques stored in the index, allowing the system to achieve high reliability in each specific region while maintaining versatility across multiple regions through the comprehensive index.
Solution Approach 2:
The patent implements parameter changes by dynamically adjusting localization parameters based on the current geographical region and driving context. The localization variant index stores optimized parameter sets for different regions, and the system automatically switches between these parameter sets as the vehicle moves between regions, thereby maintaining high localization reliability across diverse geographical areas without manual reconfiguration.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The system ensures high-precision localization by dynamically selecting the best localization technique and parameters for each context, enhancing navigation accuracy and reliability across diverse geographical and environmental conditions.
Implementation Method 1
The vehicle computing system integrates results of localization variants from different sensor modalities using Kalman filtering. The integration of localization variants based on Kalman filtering uses the measures of confidence values and measures of covariance values.
Data Source
AI summary
A vehicle, for example, an autonomous vehicle performs localization to determine the current location of the vehicle using different localization techniques as the vehicle drives. The localization technique used by the autonomous vehicle is selected from a localization variant index that stores mapping from a driving context to localization variant, each localization variant identifying a localization technique. The driving context may comprise information including: a geographical region in which the autonomous vehicle is driving, a speed at which the autonomous vehicle is driving, an angular velocity of the autonomous vehicle, or other information. Using an optimal localization technique in each driving context improves the accuracy of localization as well as computing efficiency of the process of localization.