Driving Behavior Classification via Sensor-Image Correlation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for determining automobile insurance premiums based on sensor data from vehicles do not provide sufficient context to differentiate between unsafe and safe driving behaviors, leading to potential misclassification of drivers and inaccurate premium calculations.
Innovation Solution
A method and system that combines sensor data with image data from imaging devices to classify driving events, providing context through correlation of sensor and image data to accurately assess driver behavior, and calculates insurance premiums based on these classifications.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If sensor data alone is used to classify driving behavior, then the system complexity is low, but the classification accuracy is insufficient due to lack of context
Solution Approach 1:
The patent combines sensor data (accelerometers, gyroscopes, GPS) with image data from cameras to create a multi-modal monitoring system. This merging of different data types provides contextual information that improves classification accuracy while maintaining manageable system complexity through integrated processing
Solution Approach 2:
The patent introduces an intermediary processing system that correlates sensor data with image data to determine context. This intermediary layer analyzes both data types together to distinguish between safe and unsafe driving behaviors, improving classification accuracy without requiring direct complex interaction between all system components
2Reliability
If sensor data is collected continuously, then the monitoring coverage is complete, but the ability to determine context and differentiate safe from unsafe behavior is insufficient
Solution Approach 1:
The patent adds a visual dimension by incorporating image data from cameras alongside traditional sensor data. This additional dimension provides contextual information about the driving environment and driver actions, enabling reliable differentiation between safe and unsafe behaviors that sensor data alone cannot distinguish
3Loss of information
If image data is added to sensor data, then the context information is improved, but the device complexity increases
Solution Approach 1:
The patent implements a multi-functional system where the same imaging devices and processing infrastructure serve multiple purposes: capturing driver behavior, recording driving environment context, and providing data for classification. This universal approach reduces overall system complexity compared to having separate dedicated systems for each function
Data Source
AI summary
Embodiments are provided for determining driving behavior. In an example method for determining driver behavior from a driving session, the method includes obtaining sensor data from a sensor during a driving session, obtaining image data from an imaging device during the driving session, and analyzing at least one of the sensor data and the image data from the driving session, to identify an event of interest. The method further includes assigning a classification to the event of interest according to a correlation between the sensor data for the event of interest and at least one image from the image data for the event of interest, the correlation based on time the sensor data and the image data is obtained during the driving session, and, wherein the at least one image for the event of interest provides the context for the event of interest.