Sparse Road Feature Mapping With Conformal Prediction Quality Scores
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles face challenges in navigating due to the sheer volume of data from sensors and traditional mapping technologies, which can limit or adversely affect navigation efficiency and accuracy.
Innovation Solution
A system utilizing cameras to generate a sparse map for autonomous vehicle navigation, incorporating a trained model to determine location indicators and quality values for road features, and distribute the map to a host vehicle navigation system.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional mapping technology is used to navigate autonomous vehicles, then navigation coverage is comprehensive, but data storage and processing burden increases significantly
Solution Approach 1:
The patent extracts only the essential navigation elements from complete map data by using a trained model to identify and store location indicators for critical road features (intersections, lane changes, merges) along with their quality values. This selective extraction reduces data volume while maintaining navigation reliability for autonomous vehicle operation.
Solution Approach 2:
The patent applies local quality by assigning different quality values to different location indicators based on their reliability and importance. High-quality indicators are stored and used for critical navigation decisions, while lower-quality indicators are filtered out. This ensures navigation accuracy is maintained at critical locations without storing all data points uniformly.
2Reliability
If complete map data is stored and processed, then navigation reliability is maintained, but processing time and computational resources increase
Solution Approach 1:
The system extracts only necessary location indicators and quality values from complete map data using a trained model, storing only essential information for autonomous navigation. This extraction process significantly reduces processing time while maintaining navigation reliability by focusing on critical road features.
Solution Approach 2:
The trained model performs preliminary analysis of map data to pre-identify and store only the location indicators that will be needed for navigation decisions. This preliminary action filters out unnecessary data before navigation occurs, reducing real-time processing time while ensuring reliability through pre-validated quality values.
3Measurement precision
If high-resolution map data is used, then location precision is improved, but data storage requirements increase
Solution Approach 1:
The patent implements local quality by storing high-precision location indicators only at critical road features where precise location information is essential for safe autonomous navigation (intersections, lane changes, merges). Other less critical areas use reduced precision, optimizing the balance between location precision and data storage requirements.
Solution Approach 2:
The system extracts and stores only the high-precision location indicators that are essential for navigation safety at critical features, rather than storing high-resolution data uniformly across all map areas. This selective extraction maintains necessary location precision while significantly reducing overall data storage requirements.
Data Source
AI summary
A system for generating a map for use in navigating a host vehicle relative to a road segment, including: at least one processor comprising circuitry and a memory, wherein the memory includes instructions that when executed by the circuitry cause the at least one processor to: receive road topography information representative of one or more features associated with the road segment; provide one or more indicators associated with the road topography information as input to a trained model, wherein the trained model is configured to: determine a location indicator for at least one map feature based on the one or more indicators associated with the road topography information; determine a quality value associated with the location indicator; and output the location indicator and the quality value; store in the map the determined location indicator for the at least one map feature; store in the map the determined quality value associated with the location indicator; and distribute the map to a host vehicle navigation system for use in navigating the host vehicle relative to the road segment.