Lane-Based Routing for Autonomous Vehicles
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current routing systems for autonomous driving vehicles lack efficiency and flexibility in finding optimal routes, as they rely on road-based guidance rather than lane-specific navigation.
Innovation Solution
A lane-based routing system that identifies navigable road segments, determines effective lane paths, and generates a lane-based reference line for autonomous driving vehicles, utilizing routing tables and sensors to select the most efficient lane path based on real-time traffic and road conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If road-based routing is used for autonomous driving vehicles, then general guidance can be provided, but efficiency and flexibility in searching for optimal routes is insufficient
Solution Approach 1:
The patent segments the routing system from road-level to lane-level granularity. Instead of treating roads as single entities, the system divides them into individual lane segments with unique identifiers, allowing the autonomous vehicle to plan and navigate at the lane level. This segmentation enables more precise route optimization and flexible maneuvering while maintaining computational efficiency through hierarchical processing.
2Productivity
If lane-based routing with effective distance calculation is implemented, then navigation efficiency improves, but system complexity increases
Solution Approach 1:
The system performs preliminary actions by pre-calculating and storing lane connectivity information, lane attributes, and effective distance metrics in routing tables before actual navigation occurs. This pre-processing approach allows the routing algorithm to quickly query and compare pre-computed lane paths without performing complex calculations in real-time, thereby improving navigation efficiency while managing system complexity through offline preparation.
Solution Approach 2:
The patent introduces an intermediary layer of lane-level abstraction between the road network and the vehicle navigation system. This intermediary lane segment structure acts as a mediator that simplifies complex road intersections and merges into manageable lane sequences, making the routing problem more tractable while enabling efficient pathfinding through structured lane connectivity data.
Data Source
AI summary
Embodiments disclose systems and methods to generate a lane-based reference line for an autonomous driving vehicle. In one embodiment, a system receives a request to route an autonomous driving vehicle (ADV) from a source location to a target location. In response to the request, the system identifies a set of road segments between the source location and the target location navigable by the ADV, where each of the road segments in the set includes one or more lane segments. The system identifies one or more lane paths corresponding to the set of road segments. The system determines an effective distance for each of the one or more lane paths. The system selects a lane path based on the determined effective distance. The system generates a lane-based reference line based on the selected lane path.