Vehicle Trajectory Sampling for Scalable Lane Geometry Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for generating geospatial lane data are either too costly and time-consuming due to manual curation or limited by the availability of expensive sensor-equipped vehicles, making it difficult to achieve accurate and scalable generation of lane data for large geographic regions.
Innovation Solution
Utilizing vehicle trajectory data derived from various sensors to infer the geospatial geometry of lanes by analyzing vehicle trajectories within a geographic region, allowing for scalable and accurate generation of lane data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual curation methods are used to generate geospatial lane data, then measurement precision is improved, but productivity deteriorates due to being too costly and time-consuming
Solution Approach 1:
The patent creates virtual copies of real vehicle trajectories through simulation, generating synthetic trajectory data that mimics real-world driving patterns. This allows large-scale generation of lane data without manual curation, resolving the contradiction between accuracy and productivity by using simulated copies instead of expensive manual processes
Solution Approach 2:
The system enables lane data generation to serve itself by using generated lane data to improve trajectory prediction, which in turn generates better lane data. This self-reinforcing cycle eliminates the need for continuous manual curation while maintaining improving accuracy, resolving the productivity-precision contradiction
2Measurement precision
If expensive sensor-equipped vehicles are used to collect trajectory data, then measurement precision is improved, but productivity deteriorates due to limited availability
Solution Approach 1:
The patent creates virtual copies of expensive sensor-equipped vehicles through simulation environments. These virtual vehicles can be deployed in unlimited numbers across diverse geographic regions, eliminating the scalability limitations of physical sensor-equipped vehicles while maintaining data quality through realistic simulation parameters
Solution Approach 2:
The system changes the state of vehicle data collection from physical to virtual by adjusting simulation parameters such as vehicle dynamics, sensor characteristics, and environmental conditions. This allows maintaining the quality characteristics of expensive sensor data while achieving unlimited scalability through parameter-based virtualization
3Productivity
If trajectory data from multiple geographic regions is aggregated, then productivity is improved through scalable data generation, but measurement precision deteriorates due to heterogeneity of data
Solution Approach 1:
The patent applies local quality by training separate trajectory prediction models for different geographic regions, allowing each model to capture local driving patterns and characteristics. This regional customization maintains data consistency and precision within each region while still achieving scalable productivity through the modular multi-region architecture
Solution Approach 2:
The system segments the geographic region into multiple zones with distinct traffic patterns and characteristics. By processing and generating lane data separately for each segment while maintaining a unified overall structure, the system achieves both scalability across regions and precision within each segment
Data Source
AI summary
Examples disclosed herein may involve a computing system that is operable to (i) identify a set of vehicle trajectories that are associated with a segment of a road network, (ii) identify a first cluster of sampling points between the identified set of vehicle trajectories and a first sampling position along the segment, wherein the first cluster has an associated geospatial position and is inferred to be associated with one given lane of the segment, (iii) identify a subset of vehicle trajectories in the identified set that are inferred to be associated with the given lane between the first sampling position and a second sampling position along the segment, (iv) identify a second cluster of sampling points between the identified subset of vehicle trajectories and the second sampling position, wherein the second cluster has an associated geospatial position, and (v) determine a geospatial geometry of the given lane.