Geolocation Clustering by Travel Mode for Traffic Flow Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for analyzing geolocation data from mobile communication devices struggle to accurately determine travel modes and traffic flows, particularly in scenarios with limited geolocation data points, such as commutes, leading to incomplete traffic analysis.
Innovation Solution
A method and system that cluster geolocation data using different algorithms based on travel modes (highway, light rail, footpath) and infer travel routes by analyzing cell site identities, geofenced areas, and speed thresholds, with an analysis application determining travel routes and intersections with points-of-interest to assess traffic flow.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If geolocation data is analyzed without travel mode classification, then analysis simplicity is maintained, but measurement precision of traffic flow determination deteriorates
Solution Approach 1:
The patent segments geolocation data into distinct travel modes (highway, light rail, footpath) based on cell site identity patterns, geofenced area memberships, and speed thresholds. This segmentation enables precise traffic flow determination for each mode separately, resolving the contradiction by organizing complex data into manageable categories that improve measurement precision without overwhelming processing complexity.
Solution Approach 2:
The patent changes parameters such as cell site identity, geofenced area membership, and speed threshold to classify geolocation data into different travel modes. By dynamically adjusting these parameters based on observed patterns, the system achieves accurate traffic flow measurement while maintaining systematic processing that doesn't excessively increase complexity.
2Manufacturing precision
If clustering algorithms are applied to all geolocation data uniformly, then processing simplicity is maintained, but manufacturing precision of travel route determination deteriorates
Solution Approach 1:
The patent applies different clustering algorithms to different travel mode segments rather than uniformly to all data. Highway travel uses one clustering approach, light rail uses another, and footpath uses a third, each optimized for its specific characteristics. This segmented approach improves travel route determination accuracy while keeping individual algorithm complexities manageable.
Solution Approach 2:
The patent implements local quality by tailoring clustering algorithm parameters and types to specific travel modes and geographic regions. Each cluster is optimized for its local characteristics (e.g., urban vs. rural, different transit systems), improving overall route determination precision without requiring a single overly complex universal algorithm.
3Reliability
If geolocation data points are sparse (e.g., during commutes), then data collection simplicity is maintained, but reliability of traffic analysis deteriorates
Solution Approach 1:
The patent performs preliminary actions by classifying geolocation data into travel modes and creating clusters before full traffic analysis is conducted. This pre-processing organizes sparse data into structured groups, enabling reliable traffic analysis even when individual data points are limited, as the clustering compensates for data sparsity.
Solution Approach 2:
The patent creates representative copies or proxies for sparse geolocation data through clustering, where each cluster represents multiple potential locations. This copying approach allows the system to infer traffic patterns from limited actual data points, maintaining analysis reliability without requiring abundant raw geolocation data.
Data Source
AI summary
A geolocating method to determine a traffic flow at a point-of-interest (POI). The method comprises clustering geolocation data associated with a plurality of mobile communication devices by an analysis application executing on a computer system based on different clustering algorithms associated with different travel modes identified by the geolocation data, for each of the plurality of mobile communication devices, determining travel routes traversed by the mobile communication device by the analysis application based on the clustering of the geolocation data and based on a map of travel routes, where each of the travel routes is one of a highway travel route, a light rail travel route, or a footpath travel route, and, for each of a plurality of POIs, determining by the analysis application a number of different mobile communication devices that intersect with the POI based on the travel routes traversed by the mobile communication devices.