Road Element Data Layer Augmentation for HD-Map-Free Localization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current localization solutions for autonomous vehicles rely on high-definition maps with predetermined landmarks, which are not always available and require extensive data collection, making them inefficient and less secure.
Innovation Solution
A method utilizing zero-shot learning to generate static road element information from real-time vehicle sensing, combining it with probabilistic location information and aerial maps to provide accurate vehicle localization without relying on pre-defined high-definition maps.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If high-definition maps with predetermined landmarks are used for localization, then location accuracy is improved, but system complexity and data collection requirements increase
Solution Approach 1:
The patent creates a simplified representation (copy) of the environment using neural network-generated maps instead of requiring complex high-definition maps with predetermined landmarks. This copy contains essential localization information while being much simpler to generate and update.
Solution Approach 2:
The patent replaces the traditional mechanical approach of using pre-defined map data structures with a neural network-based system that generates location representations dynamically. This substitution eliminates the need for manual map creation and landmark definition.
2Measurement precision
If high-definition maps with predetermined landmarks are used for localization, then location accuracy is improved, but data collection and processing requirements increase
Solution Approach 1:
The patent extracts only the essential information needed for localization from the environment using neural networks, rather than collecting and processing complete high-definition map data. This extraction approach retains location accuracy while dramatically reducing data volume requirements.
Solution Approach 2:
The patent changes the parameters of environmental data representation through neural network processing, transforming raw sensor data into compressed, essential feature representations. This parameter transformation maintains localization precision while reducing the quantity of data needed.
3Measurement precision
If high-definition maps are used for localization, then location accuracy is improved, but update frequency and responsiveness decrease
Solution Approach 1:
The patent implements a dynamic mapping system where the neural network continuously processes new sensor data and updates the environmental representation in real-time. This dynamic approach allows the system to adapt to changing conditions without the delays associated with updating static high-definition maps.
Solution Approach 2:
The neural network-based system maintains continuous processing of sensor data to update location representations, ensuring that the map information is always current. This continuous action eliminates the update delays inherent in traditional map-based systems.
4Measurement precision
If traditional map-based localization is used, then location determination is improved, but security and reliability are reduced
Solution Approach 1:
The system uses the vehicle's own sensor data and neural network processing to generate location information independently, rather than relying on external map providers. This self-service approach enhances security by eliminating dependencies on third-party map data that could be compromised.
Solution Approach 2:
The neural network system continuously receives feedback from sensor data and adjusts its location representations accordingly. This feedback mechanism improves reliability by allowing the system to correct errors and adapt to real-time conditions, whereas traditional maps cannot be dynamically corrected.
Data Source
AI summary
A method that is computer implemented and is for data layer augmentation, the method includes obtaining, by a processor associated with a vehicle, a data layer associated with road elements of a specified type; obtaining, by a processor associated with a vehicle, localization information regarding a location of the vehicle, wherein the road element information is obtained based on aerial image information within a region of a vehicle and on environmental information sensed by the vehicle; and augmenting the data layer using the localization information, wherein the augmenting of the data layer comprises populating a database with data representing updated road elements location for a group of road elements of the specified type within the region of the vehicle.