Virtual Lane Generation for Autonomous Vehicle Intersection Navigation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles (AVs) face challenges in identifying lane markings at road intersections, leading to potential accidents and the need to switch driving modes at traffic signal intersections, which compromises safety for both AVs and nearby vehicles.
Innovation Solution
A system and method for generating virtual lanes using processor-determined paths, lane shape points, and sensor data to guide AVs safely through intersections, eliminating the need for mode changes by providing a predefined trajectory within the virtual lane.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If autonomous vehicles rely on sensor data and lane markings for navigation, then they can operate autonomously, but they fail to identify lane markings at road intersections leading to accidents
Solution Approach 1:
The patent introduces virtual lanes as an intermediary computational construct between physical lane markings and autonomous vehicle navigation. These virtual lanes are generated by processing historical sensor data and map matching to create reliable path guidance when physical lane markings are undetectable at intersections
Solution Approach 2:
The system performs preliminary processing of historical sensor data from multiple vehicles to generate virtual lane information before the autonomous vehicle reaches the intersection. This advance preparation ensures navigation guidance is available even when real-time lane marking detection fails
2Reliability
If AVs switch driving mode at traffic signal intersections to ensure safety, then safety concerns are addressed, but continuous autonomous operation is compromised
Solution Approach 1:
The system enables autonomous vehicles to self-navigate through intersections using generated virtual lane information, eliminating the need for mode switching to manual or supervised control. The virtual lanes provide self-sufficient guidance that maintains continuous autonomous operation while ensuring safety
3Measurement precision
If virtual lanes are generated using historical sensor data and map matching, then accurate paths are provided for AVs, but data processing complexity increases
Solution Approach 1:
The patent combines historical sensor data from multiple vehicles with map data in an integrated processing system. By merging these data sources and applying map matching algorithms, the system generates accurate virtual lane information that compensates for the complexity of processing multiple data streams
Data Source
AI summary
The disclosure provides a system and a method for virtual lane generation for a vehicle. The system may determine one or more paths associated with a lane pair of the vehicle. The lane pair may correspond to an ingress lane and an egress lane associated with a traffic signal intersection. The system may further aggregate the determined one or more paths corresponding to the lane pair, to generate a bounding path associated with the lane pair. The system may further generate a virtual lane based on the generated bounding path. The generation of the virtual lane may be based on determination of a plurality of lane shape points corresponding to the generated bounding path. The system may further store data associated with the generated virtual lane in a map database.