Intersection Path Prediction Using Map Messages for Collision Warnings
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional vehicle path prediction systems are inaccurate, especially at intersections, due to varying turning radii and lack of intersection-specific path information in map messages, making it difficult to predict potential collisions with vulnerable road users.
Innovation Solution
A remote station that updates baseline static paths using road obstruction information and dynamic path data, broadcasting map messages with intersection-specific path information to vehicles, enabling accurate path prediction and collision warnings.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional vehicle path prediction systems are used, then the system complexity is low, but the path prediction accuracy is poor at intersections
Solution Approach 1:
The system pre-calculates and stores baseline static paths for common intersection maneuvers (left turn, right turn, straight) before vehicles arrive. These pre-computed paths are stored in the remote station and retrieved when vehicles approach the intersection, eliminating the need for complex real-time calculations and improving prediction accuracy.
Solution Approach 2:
A remote station is introduced as an intermediary between map data sources and vehicle systems. This remote station pre-processes map data to generate baseline static paths, receives real-time vehicle position information, and provides updated path predictions. This intermediary handles the computational complexity centrally, allowing individual vehicles to use simpler prediction algorithms while achieving high accuracy.
2Reliability
If baseline static paths are updated with road obstruction information, then the path prediction reliability is improved, but the data processing time increases
Solution Approach 1:
The system pre-processes road obstruction information and updates baseline static paths in advance before vehicles reach the intersection. The remote station maintains an up-to-date database of obstruction-free paths by continuously monitoring road conditions and pre-calculating alternative routes, so that when vehicles arrive, the most reliable paths are already prepared and can be quickly retrieved.
Solution Approach 2:
The system updates path information at predetermined frequencies and for specific high-priority vehicles approaching the intersection, rather than continuously updating for all vehicles. This partial updating approach maintains reliability for critical path predictions while reducing overall data processing time and computational burden.
3Measurement precision
If intersection-specific path information is added to map messages, then the collision warning capability is improved, but the message data volume increases
Solution Approach 1:
The system adds intersection-specific path information selectively to map messages only when vehicles are approaching or within the intersection zone. The map messages include detailed nodal points and path geometry only for the immediate intersection area, while other regions use standard map data. This localized enhancement provides collision warning capability where needed without significantly increasing overall message data volume.
Solution Approach 2:
The path information in map messages is segmented into discrete nodal points representing key locations (entry points, exit points, turning points) rather than continuous geometric data. This segmentation reduces data volume while maintaining sufficient precision for collision warning, as vehicles can interpolate positions between nodal points without requiring detailed continuous path descriptions.
Data Source
AI summary
A path prediction system includes a transceiver, a memory and a control module. The transceiver is configured to receive a map message at a host vehicle, the map message comprising path information of vehicles through an intersection. The memory is configured to store map data including global navigation satellite system information. The control module is configured to (i) determine if the host vehicle is approaching, at or in an intersection based on the map data, (ii) in response to determining that the host vehicle is approaching, at or in the intersection, predict a path of the host vehicle through the intersection based on the path information, and (iii) based on the predicted path of the host vehicle, perform at least one collision warning operation.