Virtual Lane Biasing Around Static Objects for Autonomous Navigation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles struggle to navigate safely around static objects in environments without clear lane demarcations, often driving unnecessarily close to these objects or stopping due to misidentification, despite having sufficient space to maintain a safer distance.
Innovation Solution
Systems and methods generate lane geometries using static object locations and drivable space analysis to determine new paths that avoid static objects by shifting centerlines, incorporating obstacle perception to differentiate between parked and moving vehicles, and using temporal smoothing to ensure smooth navigation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the autonomous system identifies the entire street as a single lane and drives down the middle, then the vehicle can navigate without stopping, but the vehicle may come too close to static objects like parked vehicles and bicycles
Solution Approach 1:
The system segments the street into multiple virtual lanes using generated lane markings, even in the absence of physical markings. This segmentation allows the vehicle to identify specific lanes for travel while maintaining awareness of static objects in adjacent areas, thus navigating continuously without stopping while preserving safe distances from parked vehicles and bicycles.
Solution Approach 2:
The system introduces generated lane markings as an intermediary between the vehicle and static objects. These virtual lane markings create a conceptual boundary that guides the vehicle's path while implicitly defining safe zones around static objects, allowing the vehicle to maintain productivity without compromising safety.
2Ease of operation
If the autonomous system identifies a lane as one side of the street, then the vehicle can navigate with more space from the center, but the vehicle may still navigate too closely to static objects on that side
Solution Approach 1:
The system applies local quality by generating lane markings and safety zones specifically around static objects based on their individual locations and characteristics. Rather than applying a uniform lateral offset, the system creates localized protective zones around parked vehicles, bicycles, and other static objects, allowing flexible lateral navigation while maintaining appropriate safety distances wherever needed.
3Reliability
If the autonomous system comes to a stop to avoid static objects, then collision risk is reduced, but navigation efficiency and speed are decreased
Solution Approach 1:
The system performs preliminary action by pre-generating lane markings and identifying safe travel paths before the vehicle reaches static objects. By proactively planning the navigation route around parked vehicles and bicycles using the generated lane geometry, the system eliminates the need for reactive stopping, thus maintaining high navigation speed while ensuring collision avoidance.
Data Source
AI summary
In various examples, lane biasing for navigating around objects in autonomous systems and applications is described. Systems and methods are disclosed that generate lane (or other demarcated regions of an environment) geometries in environments—such as environments without clear lane or boundary demarcations—by using locations and poses of static objects and/or outputs of a drivable free-space analysis. The systems and methods then use the lane geometries, the locations of the static objects, and current paths (e.g., centerlines of the current paths) along the lanes to determine new paths for navigating around the static objects. For instance, the new paths may be determined by shifting the centerlines of the current paths in directions away from the static objects by some distance or safety margin. This way, the vehicles may navigate around the static objects with greater distances between the vehicles and the static objects for increased safety.