Lane-Centric Road Model for Complex Intersection Navigation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing road network models fail to accurately describe complex road topologies, particularly at intersections where lane boundaries are not aligned, which is crucial for autonomous driving systems that require precise lane-level information.
Innovation Solution
A lane-centric road network model that details individual lanes with longitudinal and lateral boundary elements, along with adjacency information to define crossing parameters, allowing for accurate representation and navigation of road segments with non-trivial topologies.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a road segment-based model is used to describe road networks, then the model structure is simple and easy to implement, but the model cannot accurately describe complex road topologies with non-trivial lane level configurations
Solution Approach 1:
The patent segments the road network into discrete lane objects, where each lane is represented as an independent entity with its own boundary elements (lines and arcs). This segmentation allows precise representation of complex lane topologies while maintaining a structured data model. The lane-centric approach divides the continuous road network into manageable lane segments that can be individually processed and stored.
Solution Approach 2:
The patent introduces a new dimensional perspective by representing lanes in terms of boundary elements (one-dimensional lines and arcs) that define two-dimensional lane areas. This dimensional transformation enables accurate topological representation of complex intersections and lane configurations while keeping the underlying data structure organized and computable.
2Measurement precision
If a virtual amalgamate stop line is used to represent the stop line for an entire road segment, then the data processing is simplified, but the location accuracy for autonomous vehicle navigation is reduced
Solution Approach 1:
The patent applies local quality by assigning specific stop line positions to individual lanes rather than using a single virtual stop line for the entire road segment. Each lane can have its own stop line location, allowing the system to account for variations in stop line positions across different lanes (e.g., left turn lanes with offset stop lines). This localized approach maintains high positioning accuracy while preserving the structured lane-based data model.
3Measurement precision
If detailed lane-level boundary information is stored for each lane, then the navigation accuracy is improved, but the data storage requirements and processing complexity increase
Solution Approach 1:
The patent segments lane boundaries into discrete line and arc elements, storing only the necessary geometric parameters for each segment. This segmentation reduces data redundancy by representing continuous boundaries as a series of manageable geometric primitives, thereby decreasing overall data storage requirements while maintaining precise lane boundary representation.
Solution Approach 2:
Instead of storing lane areas directly, the patent inverts the approach by storing the boundary elements (lines and arcs) that define the lanes. This inversion reduces data storage needs because boundaries can be represented more compactly than full area definitions, while still enabling precise reconstruction of lane geometries for navigation purposes.
Data Source
Figure 1
Figure 2A
Figure 2B
AI summary
A geographic database storing map data is provided. The geographic database is stored in a non-transitory computer readable medium. The geographic database comprises a plurality of records corresponding to drivable surfaces of a road network. The plurality of records comprise a plurality of lane records corresponding to particular lanes of the road network. Each first record of the plurality of records comprises a plurality of instances of adjacency information. Each instance of adjacency information/data (a) links the first record corresponding to a first drivable surface of the road network to a second record of the plurality of records corresponding to a second drivable surface of the road network. The first drivable surface is adjacent to the second drivable surface. Each instance of adjacency information/data indicates crossing parameters between the first drivable surface and the second drivable surface.