Lane-Specific Speed Patterns from Probe Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional digital maps rely on average traffic speed data, which fails to accurately represent lane-specific traffic conditions, leading to insufficient detail and inaccurate understanding of traffic patterns along road segments.
Innovation Solution
A method to analyze probe data from multiple sources to establish a multi-modal distribution of vehicle speeds along road segments, allowing for the generation of lane-specific speed patterns by matching probe data to individual lanes based on distance from a reference position, and considering factors like time and weather conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If average traffic speed data is used for all vehicles on a road segment, then the system complexity is reduced and data processing is simplified, but the accuracy and detail of lane-specific traffic representation deteriorates
Solution Approach 1:
The patent segments the road segment into multiple lanes and further segments the traffic data into lane-specific groups. By dividing the probe data according to lane positions (using distance from reference position and multi-modal distribution analysis), the system processes each lane separately, thereby improving measurement precision for lane-specific traffic while maintaining manageable complexity through systematic segmentation
Solution Approach 2:
The patent applies local quality by providing different speed representations for different lanes rather than a uniform average for the entire road segment. Each lane receives its own speed pattern based on local probe data analysis, allowing the system to capture lane-specific traffic characteristics (e.g., slower speeds in right lanes near exits, faster speeds in left lanes) without overwhelming complexity
2Productivity
If probe data is analyzed at the road segment level only, then the data processing is simpler and faster, but the detail and understanding of individual lane traffic patterns is insufficient
Solution Approach 1:
The patent segments probe data by lane using distance calculations from reference positions and identifies multi-modal distributions to separate traffic into lane-specific groups. This segmentation enables the system to process data efficiently at scale while preserving lane-specific information, as each lane's speed pattern is derived from its corresponding probe data subset
Solution Approach 2:
The patent introduces a spatial dimension (distance from reference position) to differentiate lanes within a road segment. By analyzing the distribution of probe data along this dimensional axis and identifying peaks in the multi-modal distribution, the system recovers lane-specific information that would otherwise be lost in aggregate road-level averaging
Data Source
AI summary
A method is provided to generate vehicle lane speed patterns. A method may include: receiving probe data from a plurality of probes, where the probe data includes probe data point location and heading; matching probe data points to a road segment to generate map-matched probe data points; analyzing the probe data relative to the road segment to establish a multi-modal distribution of probe data representing a distance of the probe data points from a pre-defined reference position of the road segment; matching the analyzed probe data points to individual lanes of the road segment, where peaks in the established multi-modal distribution are associated with individual lanes; and generating a vehicle lane speed pattern for each lane of the road segment based on a speed associated with probe data that is map-matched to the individual lanes.