Vehicle Automation Level Classification via Sensor Signatures
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Service providers face challenges in delivering value and convenience to consumers due to varying accuracy, reliability, and relevancy of sensor data from different vehicles, necessitating an approach for automated vehicle classification based on automation levels.
Innovation Solution
A method and apparatus that determine training sensor data from vehicles with known automation levels, extract sensor signatures based on classification features, and classify other vehicles accordingly, using machine learning to identify manual, partial, or fully autonomous vehicles, enabling customized services and quality scoring of probe data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If sensor data from multiple vehicles is collected and processed, then service quality and customization improve, but data accuracy and reliability vary making classification difficult
Solution Approach 1:
The patent segments vehicles into different automation levels (manual, partial automation, full automation) based on their sensor data characteristics. This segmentation allows service providers to classify and treat vehicle data differently according to automation level, improving service customization while managing data quality variations through targeted classification criteria for each segment.
Solution Approach 2:
The patent changes the classification parameters from generic vehicle attributes to automation-level-specific sensor data patterns. By adjusting classification thresholds and features based on automation level (e.g., different sensor suites for manual vs. autonomous vehicles), the system improves measurement precision for each vehicle category while maintaining versatility across diverse vehicle types.
2Productivity
If automated vehicle classification is implemented, then resource allocation and traffic safety improve, but system complexity increases
Solution Approach 1:
The patent performs preliminary classification of vehicles into automation levels before resource allocation decisions are made. By pre-categorizing vehicles based on their automation characteristics, the system simplifies subsequent resource allocation, traffic management, and safety interventions, reducing overall system complexity while improving productivity through streamlined decision-making processes.
Solution Approach 2:
The patent introduces an intermediary classification layer that translates complex sensor data into simplified automation level categories. This intermediary classification system acts as a mediator between raw sensor data and resource allocation processes, reducing complexity by converting diverse data inputs into standardized classification outputs that facilitate efficient resource management.
3Measurement precision
If sensor signatures are extracted for classification, then vehicle automation level detection accuracy improves, but data processing time and computational requirements increase
Solution Approach 1:
The patent extracts specific sensor signatures and key features from complete sensor datasets, isolating only the most discriminative characteristics needed for automation level classification. By taking out only the essential features (e.g., specific sensor patterns unique to each automation level) rather than processing entire datasets, the system improves detection accuracy while reducing computational time and resource requirements.
Data Source
AI summary
An approach is provided for classifying one or more vehicles based on their level of automation. The approach involves determining training sensor data collected during at least one driving operation of one or more vehicles, wherein one or more automation levels of the one or more vehicles are known. The approach also involves determining one or more sensor signatures for the one or more automation levels based, at least in part, on one or more values of one or more classification features extracted from the training sensor data. The approach further involves causing, at least in part, a classification of one or more other vehicles according to the one or more automation levels based, at least in part, on the one or more sensor signatures and sensor data associated with the one or more other vehicles.