Roadside Traffic Object Lane Assignment Using Driving Behavior
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for autonomous vehicles to identify and assign roadside traffic objects to specific lanes are inaccurate and complex, especially in multi-lane scenarios, relying heavily on logic-based solutions that struggle with real-world complexity.
Innovation Solution
A data-driven approach using machine learning algorithms that generate training data by correlating roadside traffic object identification with changes in vehicle behavior, allowing for accurate lane assignment through image annotation and training data sets, adaptable to varying traffic situations and infrastructure.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If logic-based solutions are used for lane assignment of traffic objects, then the system structure is simple, but the accuracy and reliability of lane assignment deteriorates in complex multi-lane scenarios
Solution Approach 1:
The patent replaces logic-based rule systems with machine learning models that process sensor data, images, and map information to determine lane assignments. The ML algorithm learns patterns from training data comprising sensor data, images, map data, and ground truth lane assignments, enabling accurate classification of traffic objects to specific lanes in complex multi-lane scenarios without relying on brittle rule-based logic.
2Ease of manufacture
If rule-based systems are used for traffic object classification, then the implementation is straightforward, but the adaptability to varying traffic situations and infrastructure deteriorates
Solution Approach 1:
The patent implements a dynamic system where the lane assignment model is trained on diverse training data representing various traffic situations, road infrastructures, and environmental conditions. The ML algorithm adapts to different scenarios by learning from labeled examples during training, allowing the system to generalize to new traffic situations and infrastructure types without requiring manual rule updates.
3Adaptability or versatility
If complex logic-based methods are used to handle multiple roads and lanes, then comprehensive coverage is achieved, but the computational complexity and processing time increases
Solution Approach 1:
The patent transforms the lane assignment problem from a complex logical reasoning task into a parameter-based machine learning classification task. By representing traffic objects, lanes, and their relationships as structured parameters in training data (sensor data, images, map data, ground truth labels), the system efficiently processes multi-lane scenarios through learned patterns rather than exhaustive logical evaluation.
Data Source
AI summary
A method, system, a vehicle and a computer-readable storage medium for lane assignment of a roadside traffic object present on a road. The method includes obtaining sensor data of an ego vehicle comprising an Automated Driving System (ADS) and traveling on the road. The method further includes identifying the roadside traffic object in the surrounding environment of the ego vehicle and determining a change in a driving behavior of the ego vehicle being present in the surrounding environment of the ego vehicle. The method further includes determining a co-occurrence of the identification of the at least one roadside traffic object and determination of the change in the driving behavior of the ego vehicle. The method further includes if the co-occurrence is determined, generating a corresponding image annotation for one or more obtained images of the at least one identified roadside traffic object.