Vehicle Positioning via Map Segmentation and Sensor Fusion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current navigation systems for highly automated driving require highly accurate geometric maps, which are not suitable for conventional navigation between locations, limiting the flexibility and precision of vehicle positioning and trajectory planning.
Innovation Solution
A method and device that allow for flexible utilization of map data by determining a vehicle's position in a first map and switching to a second map segment with overlapping areas, using detection data from environment sensors, enabling modular and cost-effective map generation and use of different data collection methods.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If highly accurate geometric maps are used for automated driving, then positioning accuracy and trajectory planning are improved, but map storage requirements and system complexity increase
Solution Approach 1:
The highly accurate geometric map is divided into multiple map segments, each covering a specific geographic region. The system only loads and processes the map segment relevant to the vehicle's current location, rather than managing the entire high-precision map at once. This segmentation reduces memory requirements and processing complexity while maintaining positioning accuracy in the active region.
Solution Approach 2:
The system applies different map data qualities to different spatial regions: high-precision geometric map data is used only in the current local area where the vehicle is operating, while lower-precision or no map data is used in distant regions. This local quality approach ensures positioning accuracy where needed without the complexity and storage burden of maintaining high precision everywhere.
2Manufacturing precision
If highly accurate geometric maps are used for automated driving, then trajectory planning is improved, but storage requirements increase
Solution Approach 1:
The complete high-precision map is segmented into multiple smaller map segments. The system loads only the current map segment into memory for trajectory planning, discarding or de-loading other segments. This maintains trajectory planning precision in the active region while significantly reducing storage and memory requirements.
Solution Approach 2:
The system pre-loads map segments that are anticipated to be needed based on the planned travel route, while keeping only the current segment fully loaded in memory. This preliminary action ensures trajectory planning precision is maintained for upcoming regions without requiring all map data to be simultaneously stored in high precision.
3Device complexity
If a single map system is used for both navigation and automated driving, then system simplicity is improved, but positioning accuracy for automated driving deteriorates
Solution Approach 1:
The system merges two previously separate map systems (navigation map and high-precision geometric map) into a unified map segment system. The same map segment structure serves both navigation routing functions and automated driving positioning functions, eliminating the need to maintain and switch between separate map systems while preserving the positioning accuracy required for automated driving.
Solution Approach 2:
The map segment data structure is designed to be universal, serving multiple functions: it provides routing and navigation information for conventional navigation, and simultaneously provides the high-precision geometric reference needed for automated driving positioning and trajectory planning. This multi-functionality eliminates the need for separate specialized maps.
4Quantity of substance
If conventional navigation maps are used, then storage requirements are reduced, but positioning accuracy for automated driving deteriorates
Solution Approach 1:
The high-precision geometric map is segmented so that only the current local segment is loaded into memory at any time. This allows the system to use detailed geometric map data for positioning accuracy without requiring the entire high-precision map to be stored simultaneously, thus reducing memory requirements while maintaining positioning accuracy.
Solution Approach 2:
The system uses high-quality geometric map data locally in the current map segment for accurate positioning, while relying on lower-quality navigation map data for distant regions and routing. This local quality approach ensures positioning accuracy where the vehicle is operating without the storage burden of maintaining high precision everywhere.
Data Source
AI summary
As a function of a provided geoposition for the vehicle, a first position of the vehicle is determined in a first map, which represents a first geographic region. As a function of the determined first position of the vehicle in the first map, a map segment of a second map is determined. The second map has several map segments, which each represent predefined segments of a predefined second geographical region, and wherein directly adjoining segments overlap one another in a predefined manner. As a function of detection data of at least one predefined environment sensor of the vehicle and as a function of second map data of the determined map segment, a second position of the vehicle is determined in the map segment.