HD Map Data Selection for Fast Landmark Updates
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing high-definition maps require accurate and timely updates to include landmarks with decimeter precision, yet current methods struggle to efficiently process and communicate large volumes of data at high rates for vehicle navigation and control.
Innovation Solution
A system and method for determining a set of information to produce a map using vehicle positions and camera data, optimizing data characteristics and aspects to be included, allowing for rapid production and transmission of high-definition maps with landmarks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If high-definition maps include additional layers of information (geometric map, semantic map) to improve accuracy within decimeter, then manufacturing precision improves, but device complexity increases
Solution Approach 1:
The HD map is segmented into multiple functional layers: base map layer for general geographic information, geometric map layer for precise spatial relationships and drivable surfaces, and semantic map layer for object meanings and relationships. This segmentation allows each layer to be optimized independently for its specific purpose while maintaining overall map accuracy within decimeter precision.
Solution Approach 2:
The patent transitions from two-dimensional conventional maps to three-dimensional HD maps by adding vertical dimension information through voxelized geometric representations. This enables accurate representation of multi-story buildings, underground structures, and elevated roadways, improving manufacturing precision for automated vehicle control systems that require spatial awareness beyond flat surfaces.
2Reliability
If HD maps are updated at high rate to account for changes in objects or positions, then reliability improves, but loss of time increases
Solution Approach 1:
The system extracts and processes only the essential information needed for each map layer independently. The base map layer captures static geographic features, the geometric map layer extracts spatial relationships and drivable surfaces, and the semantic map layer extracts object meanings. This extraction approach reduces processing time while maintaining reliability by updating only the necessary data elements at high rates to reflect changes in objects or positions.
Solution Approach 2:
The patent implements preliminary processing of map data by pre-organizing information into structured layers with defined schemas and data models. This preliminary organization enables faster updates when changes occur, as the system can efficiently locate and modify specific layers without reprocessing entire map datasets, thereby improving reliability while reducing time loss.
3Measurement precision
If HD maps use SLAM technique with proprioception information and perceptual information to estimate and correct pose, then measurement precision improves, but device complexity increases
Solution Approach 1:
The system merges multiple sensor inputs including proprioception information from GNSS receivers, IMUs, and odometry systems with perceptual information from lidar, cameras, and other sensors. This combination enables accurate pose estimation and correction through SLAM techniques, achieving measurement precision within decimeter accuracy while managing device complexity through integrated sensor fusion architecture.
Data Source
AI summary
A system for determining a set of information to be used to produce a map of a region can include a processor and a memory. The memory can store a set determination module, a production module, and a communications module. The set determination module can determine, from: (1) positions of a vehicle that traversed a region and (2) data affiliated with images produced by a camera on the vehicle, a set of information to be used to produce a map of the region. An optimal state of the set can be based on one or more of: (1) a characteristic of the information or (2) an aspect, in the region, to be included in the map. The production module can produce, from the set, the map. The communications module can transmit the map to a specific vehicle to control movement of the specific vehicle.