Vehicle Sensor Data Synchronization Using Turn Event Alignment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for synchronizing sensor data from multiple systems, especially when used offline, face challenges in matching synchronization points between different types of data, such as images and acceleration data, leading to inconsistencies in data analysis, particularly in autonomous driving applications.
Innovation Solution
A data synchronization device comprising a first estimation unit for determining the data range of vehicle motion in camera images and a second estimation unit for determining the data range of vehicle motion in acceleration data, with a synchronization unit performing time-synchronization based on these estimated ranges to align time-series data of images and acceleration data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If sensor data is recorded and used offline later, then data can be retrieved and analyzed at convenience, but the data cannot be consistent with actual events that occurred
Solution Approach 1:
The patent applies preliminary action by attaching time stamps to sensor data at the moment of acquisition before offline use. This ensures that when data is retrieved later, the time synchronization information is already embedded, allowing consistent correlation with actual events without requiring real-time processing.
2Measurement precision
If synchronization is performed by attaching sensors to the same system, then time synchronization is achieved, but synchronization gaps occur during system operation requiring periodic synchronization
Solution Approach 1:
The patent uses a common time reference or synchronization signal as an intermediary between multiple sensors and systems. This mediator enables continuous time synchronization without requiring periodic manual intervention, as all systems reference the same time source independently.
3Measurement precision
If different types of sensor data are synchronized by matching data with the same meaning, then synchronization points can be detected, but this method cannot be applied when data types differ significantly
Solution Approach 1:
The patent implements a universal time stamping mechanism that works across all sensor data types regardless of their specific characteristics. By assigning time metadata to all data types uniformly, the system achieves synchronization capability that is adaptable to images, acceleration data, and other diverse sensor outputs without requiring type-specific matching logic.
Data Source
AI summary
In a data synchronization device (20), a first estimation unit (51) estimates as a first data range, a data range in which a vehicle is turning right or left, in moving image data (33). A second estimation unit (52) estimates as a second data range, a data range in which the vehicle is turning right or left, in acceleration data (34). A synchronization unit (53) performs a time-synchronization between the moving image data (33) and the acceleration data (34) based on a time range corresponding to the first data range estimated by the first estimation unit (51) and a time range corresponding to the second data range estimated by the second estimation unit (52).