Vehicle Route Crossing Prediction With Yield Control Feedback
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles face challenges in accurately predicting and controlling object crossings, leading to potential safety issues due to false positive or negative detections of objects crossing their routes.
Innovation Solution
The system uses machine learning models to predict object trajectories and determine vehicle controls based on timestamp information, filtering out false positives and negatives by comparing predicted and ground truth trajectories, and adjusting vehicle movement to avoid collisions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If machine learning models are used to predict object trajectories, then object detection accuracy is improved, but false positive and false negative detections still occur
Solution Approach 1:
The system implements feedback by comparing predicted trajectories with ground truth trajectories to identify false positives and false negatives. This feedback loop enables the system to learn from errors and improve detection reliability over time, directly addressing the contradiction between detection accuracy and false positive/negative rates.
Solution Approach 2:
The system performs preliminary filtering of trajectory predictions by comparing them against ground truth data before final decision-making. This preliminary action allows the system to eliminate obvious false positives and negatives early in the processing pipeline, improving overall reliability without sacrificing detection accuracy.
2Reliability
If the vehicle adjusts movement to avoid collisions, then safety is improved, but vehicle productivity and route efficiency may deteriorate
Solution Approach 1:
The system applies partial action by adjusting vehicle movement only when and where necessary to avoid collisions, rather than continuously modifying the route. This selective approach maintains safety while minimizing impact on overall route efficiency and productivity.
Solution Approach 2:
The system dynamically adjusts vehicle control based on real-time trajectory predictions and actual object behavior. This dynamic adaptation allows the vehicle to maintain high productivity when the path is clear while ensuring safety when potential collisions are detected, resolving the contradiction between safety and efficiency.
Data Source
AI summary
Provided are methods, systems and computer program products for predicting vehicle crossing and yielding, which can include receiving sensor information indicating an object surrounding a vehicle. Some methods also include determining a future position of the vehicle based on a first trajectory of the vehicle, determining a future position of the object based on a second trajectory of the object, and determining a vehicle control based on the future position of the vehicle and the future position of the object. The methods also include training a model using the vehicle control, the first trajectory of the vehicle, and the second trajectory of the object.