Road Database Accuracy via Vehicle Trajectory Aggregation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current navigation databases, such as the TIGER database, are labor-intensive to develop and lack accuracy in representing road connectivity and drivability, leading to nonsensical navigation instructions due to positional inaccuracies and missing features like grade separations and one-way roads.
Innovation Solution
A method to generate an improved road network database by collecting and processing vehicle position data from vehicles traveling on the roads, categorizing data points, and aggregating them to create a semi-continuous trace of road geometry, which is then compared to existing databases to enhance accuracy and completeness.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional methods (aerial photographs, human observers) are used to develop navigation databases, then database coverage can be achieved, but the development process becomes labor intensive with substantial costs and limitations in accuracy
Solution Approach 1:
The system uses vehicles already traveling on roads to automatically collect position and heading data, eliminating the need for dedicated data collection missions. The vehicles serve themselves as data collection platforms, continuously contributing to database refinement without requiring special intervention or resources.
Solution Approach 2:
The patent replaces manual field verification and aerial photograph interpretation with automated electronic data collection from vehicle-mounted sensors. GPS receivers and heading sensors automatically capture position and orientation data, which is then processed through algorithms to refine road geometry, substituting mechanical human labor with electronic automation.
2Reliability
If existing databases like TIGER are used without modification, then database availability is maintained, but absolute positional accuracy and fine-scale road connectivity representation are mediocre
Solution Approach 1:
The system continuously compares newly collected vehicle position and heading data against the existing database, identifying discrepancies and systematically refining the database. This feedback loop progressively improves road position accuracy and connectivity representation by incorporating real-world observation data from actual vehicle trajectories.
Solution Approach 2:
The system performs preliminary data collection and processing to identify and correct database errors before they affect navigation operations. By proactively refining the database using accumulated vehicle data, the system prevents propagation of positional inaccuracies and connectivity errors into navigation routes.
3Manufacturing precision
If detailed field verification is performed to improve database accuracy, then road geometry precision improves, but development time and costs increase substantially
Solution Approach 1:
The system continuously collects and processes vehicle data in real-world operating conditions, rather than performing discrete field verification campaigns. This continuous data accumulation and processing refines road geometry accuracy progressively over time, eliminating the need for time-consuming periodic field surveys while maintaining improving precision.
Data Source
AI summary
Methods and systems for generating, deriving, and enhancing drivable road databases are provided. A baseline road in a road network is defined and position and/or trajectory data collected by vehicles traveling the baseline road are compiled and compared to a representation of the baseline road in an existing database. Identity and/or other property information about the road are assigned form the existing database to the new database.