Asynchronous Sensor Data Synchronization Using Interpolation and Sampling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing sensors collect data at different rates, leading to asynchronous data sources that are difficult to synchronize, which complicates analysis and reconstruction tasks.
Innovation Solution
A method and system for synchronizing data by accessing input video and telematic data at different rates, generating reduced telematic data based on iteratively-defined reference lines, interpolating data to a synchronized rate, and sampling at a threshold time period to align data points.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If data from multiple sensors is collected at different rates, then each sensor can operate at its optimal data collection rate, but the data becomes asynchronous and difficult to synchronize
Solution Approach 1:
The patent segments the data synchronization process into distinct stages: (1) collecting data at native asynchronous rates from multiple sensors, (2) identifying key data points using reference lines, (3) generating interpolated data between key points, and (4) sampling at synchronized intervals. This segmentation allows each sensor to maintain its optimal collection rate while the system as a whole achieves synchronization through structured processing of the asynchronous data streams.
Solution Approach 2:
The patent applies preliminary action by pre-defining reference lines through the asynchronous data before interpolation occurs. These reference lines are established in advance to guide the selection of key data points and to provide a framework for generating interpolated values. This preliminary structuring of the data enables efficient synchronization without requiring real-time coordination between sensors during data collection.
2Measurement precision
If all data points from high-rate sensors are retained, then data quality is maintained, but resource burden increases
Solution Approach 1:
The patent extracts only the essential data points from high-rate sensor streams by comparing actual data points against iteratively-defined reference lines. Data points that deviate significantly from the reference lines are identified as key points and retained, while points that closely follow the reference lines are represented by interpolated values. This extraction process maintains measurement precision for critical events while dramatically reducing the overall data volume that needs to be stored and processed.
Solution Approach 2:
The patent changes the parameter representation of the data by transforming continuous high-rate sensor data into a hybrid representation consisting of discrete key data points and continuous interpolated segments. This parameter change allows the system to maintain high measurement precision at key moments while using efficient interpolation for transitional periods, thereby reducing data volume without sacrificing essential information quality.
3Measurement precision
If data is synchronized at a high rate, then alignment precision is improved, but processing complexity increases
Solution Approach 1:
The patent applies partial action by synchronizing data at varying rates depending on the content and importance of the data segments. Rather than uniformly sampling all data at the maximum possible rate, the system uses interpolated sampling at lower rates for stable periods and switches to higher-rate key point capture when significant events occur. This partial application of high-rate sampling maintains alignment precision for critical data while reducing overall processing complexity during mundane time periods.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A synchronized data set is generated from multiple asynchronous data sets. An interpolated data set is generated for each asynchronous data set, at a higher data rate than recorded data rates for the asynchronous data sets. The interpolated data for each data set is then sampled at a lower data rate which is common to all the interpolated data sets. The sampled data points are synchronized between the different data sets, such that the sampled data points represent a synchronized data set. Input telematic data can be reduced to inflection points in the telematic data, and interpolation and sampling of asynchronous data sets can be limited to threshold time periods around the inflection points.