Reversible Lane Direction Detection via GNSS Probe Trajectories
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Reversible lanes pose a challenge for traffic information providers and autonomous driving systems due to the difficulty in accurately identifying probe vehicles and determining the active direction of these lanes using Global Navigation Satellite System (GNSS) data, which is plagued by noise and insufficient accuracy.
Innovation Solution
A method that involves receiving and analyzing sequences of probe data from multiple vehicles to determine the travel direction and match them to a lane based on a predetermined vehicle lane pattern, constructing probe trajectories, and using lane change probabilities to identify the active direction of reversible lanes, thereby providing accurate traffic information for navigation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If GNSS location information is used to determine probe vehicle locations, then real-time location data can be obtained, but the accuracy is insufficient to determine the particular lane due to noise and 7.8 meter confidence interval
Solution Approach 1:
The patent segments the probe data processing into multiple stages: initial lane assignment based on GNSS coordinates, trajectory construction from sequential data points, and lane determination based on trajectory analysis. This segmentation allows the system to use low-precision GNSS data for initial filtering while achieving high-precision lane-level accuracy through the multi-stage processing pipeline.
Solution Approach 2:
The patent performs preliminary actions by pre-establishing road geometry data, lane configurations, and reference trajectories before processing probe data. This preliminary preparation enables the system to quickly match incoming GNSS data to appropriate lanes without requiring complex real-time calculations, thus maintaining real-time performance while improving accuracy.
2Measurement precision
If probe data is collected from multiple sources to improve lane determination accuracy, then more comprehensive traffic information can be obtained, but the device complexity and data processing requirements increase
Solution Approach 1:
The patent implements self-service mechanisms where the system automatically filters, validates, and processes probe data from multiple sources using predefined criteria and algorithms. The trajectory analysis and lane assignment are performed autonomously without requiring manual intervention or complex coordination between multiple processing systems, thus reducing operational complexity while maintaining high accuracy.
Solution Approach 2:
The patent introduces trajectory construction as an intermediary processing layer between raw GNSS data and final lane determination. This intermediary step transforms noisy, multi-source probe data into structured trajectory representations that are easier to analyze and match against lane configurations, simplifying the overall processing architecture while improving accuracy.
3Ease of manufacture
If traditional GNSS-based lane identification is used, then the system is simple to implement, but it cannot reliably identify probe vehicles in reversible lanes due to insufficient accuracy
Solution Approach 1:
The patent applies dynamics by transitioning from static lane assignment based on single GNSS coordinates to dynamic trajectory-based lane determination. The system continuously updates lane assignments based on sequences of probe positions, allowing it to adapt to the changing nature of reversible lanes and maintain reliable identification even as traffic patterns and lane directions change.
Solution Approach 2:
The patent adds the temporal dimension to lane identification by using sequences of GNSS positions over time to construct trajectories. This transforms the problem from a two-dimensional spatial matching task to a three-dimensional spatiotemporal analysis, enabling reliable reversible lane identification while maintaining implementation feasibility through algorithmic processing.
Data Source
AI summary
In an example embodiment, a plurality of sequences of instances of probe data are received. Each sequence of instances of probe data is captured and provided by a probe apparatus comprising a plurality of sensors and is onboard a vehicle. An instance of probe data comprises location information indicating a location of the corresponding probe apparatus and the instances are ordered by capture time to form the sequence of instances. A travel direction of each probe apparatus is determined based on the corresponding sequence. Each probe apparatus is matched to a lane of a road segment based on the determined travel direction and a predetermined vehicle lane pattern. The vehicle lane pattern comprises at least one reversible lane. Probe apparatuses matched to the at least one reversible lane are identified. An active direction is determined based on the number of identified probe apparatuses corresponding to each travel direction.