Digital Street Network Database Creation from Probe Traces
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for creating digital transportation networks are limited in generating high-precision maps, especially for Advanced Driver Assistance Systems (ADAS), as they require accurate location measurements, which are not sufficiently provided by single location sensors, and cannot build networks from scratch using uncoordinated probe traces, and fail to automate the junction of transportation network segments.
Innovation Solution
The method involves acquiring sequential location measurement data from navigation devices, clustering and processing these data to identify choke points and maneuvers, and creating a digital representation of transportation networks without requiring a seed representation, allowing for on-the-fly construction and storage as polylines or mathematical curves, enabling precise vehicle routing and navigation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If uncoordinated probe traces from location sensors are used to create digital transportation networks, then the network can be built from scratch without seed representation, but the location accuracy is insufficient (±10 to 15 meters) for high-precision applications like ADAS
Solution Approach 1:
The patent combines multiple probe traces from numerous uncoordinated vehicles and pedestrians into a collective dataset. By merging these traces and identifying common paths through statistical analysis, the system achieves high-precision network creation from scratch without requiring seed representation, while compensating for individual sensor inaccuracies through crowd-sourced data aggregation.
Solution Approach 2:
The patent creates a digital copy of the transportation network by processing probe trace data into a structured representation that mirrors real-world paths. This digital model captures geometric features, maneuver points, and network segments, enabling high-precision navigation applications without requiring physical measurement equipment.
2Device complexity
If conventional location sensors are used for creating digital transportation networks, then the equipment is simple and widely available, but the positional accuracy (±10 to 15 meters) is insufficient for ADAS applications requiring less than 5 meters accuracy
Solution Approach 1:
The system uses the location sensors' inherent capabilities to collect data, then processes this data through algorithms that identify patterns and common paths. The method transforms the sensors' limitation (individual inaccuracy) into an advantage by using their ubiquity and self-collected data to build high-precision networks without requiring specialized measurement equipment.
Solution Approach 2:
The patent changes the approach from relying on single-sensor precision to using statistical parameters derived from multiple sensors. By analyzing the distribution and convergence of numerous probe traces, the system achieves sub-5-meter accuracy through parameter aggregation and pattern recognition rather than through individual sensor precision.
3Measurement precision
If probe traces are used to update existing transportation networks, then the accuracy of existing networks can be refined, but the method cannot create new networks from scratch without known segment locations
Solution Approach 1:
Instead of using known network segments to interpret probe traces (conventional approach), the patent inverts the process by using probe traces to discover and define network segments de novo. The system identifies paths, intersections, and geometric features directly from the trace data, enabling network creation from scratch while maintaining high precision through statistical analysis of the inverted data flow.
4Reliability
If manual methods are used to identify and connect transportation network segments at junctions, then the network topology can be accurately represented, but the process is time-consuming and not automated
Solution Approach 1:
The patent replaces manual identification and connection of network segments at junctions with automated algorithms that process probe trace data. The system automatically detects intersection points, identifies connecting segments, and constructs network topology through computational analysis of trace patterns, eliminating manual intervention while maintaining accurate topological representation.
Solution Approach 2:
The patent introduces maneuver points and geometric features as intermediaries between raw probe traces and final network topology. These intermediary elements serve as automated connection points that bridge multiple segments at junctions, enabling the system to automatically reconstruct complex intersection topologies without manual processing while preserving topological accuracy.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
In a method for creating a digital representation of a transportation network location measurement data is acquired, and a plurality of maneuvers between choke points are generated from the location measurement data. The digital representation of the transportation network is then created and stored based on the plurality of maneuvers.