Intersection Traversal Detection Using Telematics Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for determining intersection traversals using telematics data are processor-intensive and inefficient, as they often analyze vehicle trips that do not actually involve intersection traversals, leading to unnecessary processing power and time consumption.
Innovation Solution
A method that isolates telematics data points within an intersection by comparing them to a shape representing the intersection, then to vector lines representing entrances and exits, and finally to an average simplified line of other vehicles' paths, to determine if a vehicle turned through the intersection, thereby reducing processing power by only analyzing relevant trips.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If all vehicle trips are analyzed for intersection traversals, then comprehensive data collection is achieved, but processing time and computational resources are excessively consumed
Solution Approach 1:
The patent applies preliminary action by performing initial filtering of vehicle trips using geometric criteria (checking if trip paths intersect with intersection polygons) before conducting detailed traversal analysis. This preliminary step eliminates obviously non-relevant trips early in the processing pipeline, reducing the volume of data requiring intensive computation while ensuring no potentially relevant trips are missed.
Solution Approach 2:
The patent segments the analysis process into distinct stages: (1) preliminary geometric filtering of trips against intersection locations, (2) isolation of telematics data points within intersection boundaries, (3) comparison to vector lines for entrance/exit identification, and (4) final traversal determination. This segmentation allows each stage to process only the necessary subset of data, significantly reducing overall computational burden.
2Measurement precision
If detailed analysis is performed on all vehicle trips, then accuracy in determining intersection traversals is improved, but processing power consumption increases
Solution Approach 1:
The patent applies partial action by performing detailed traversal analysis only on vehicle trips that pass the preliminary geometric filter, rather than analyzing all trips with full detail. The filtering criterion is deliberately set to be necessary but not overly restrictive, ensuring that all potentially relevant trips are included while excluding clearly irrelevant ones, thus optimizing the balance between accuracy and computational efficiency.
3Measurement precision
If comprehensive comparison with other vehicles' trips is performed, then statistical accuracy is improved, but computational complexity increases
Solution Approach 1:
The patent segments the comparison process by first identifying trips that are relevant to each specific intersection using geometric filtering, then performing detailed comparisons only within those segmented subsets. This approach allows statistical analysis to be conducted on appropriate data subsets for each intersection, maintaining statistical accuracy while avoiding the computational complexity of comparing all trips across all intersections.
Data Source
AI summary
Techniques for using vehicle telematics data for high speed determination of intersection traversals are provided. Telematics data from a vehicle trip may be obtained and compared to an area associated with an intersection to attempt to isolate telematics data points that are within the intersection. A simplified line generated based on the isolated telematics data points is then compared to vector lines representing entrances and exits of the intersection to attempt to identify an intersection entrance and exit from the vehicle trip. The simplified line is then compared to an average simplified line representing trips of other vehicles through the same entrance and exit of the same intersection to determine whether the vehicle turned through the intersection. If it is determined that the vehicle turned through the intersection, the telematics data associated with the vehicle trip may be compared to telematics data associated with other vehicle trips through the intersection.