Time Series Forecastability Filtering for Sensor Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In manufacturing and industrial settings, practitioners face challenges in accurately forecasting time series data from thousands of sensor-emitted data streams, as existing methods are computationally expensive and time-consuming, lacking tools to eliminate unforecastable series and prioritize forecastable ones.
Innovation Solution
A method that identifies time series patterns in historical data snapshots, calculates variability measures, and determines forecastable vs. non-forecastable patterns, allowing for efficient modification of systems to generate only forecastable patterns, thereby improving system efficiency and reducing computational resources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If practitioners create thousands of forecast models using sophisticated machine learning techniques, then forecast accuracy may improve, but computational cost and time consumption increase significantly
Solution Approach 1:
The patent segments the time series data into different patterns and categories, allowing practitioners to focus on specific forecastable patterns rather than creating thousands of models for all data. This segmentation reduces computational complexity while maintaining forecast accuracy for relevant patterns.
Solution Approach 2:
The patent applies different analysis approaches to different time series patterns based on their forecastability characteristics. By identifying and focusing on patterns with lower variability and higher forecastability, the system applies sophisticated modeling only where needed, reducing overall computational burden while maintaining accuracy.
2Measurement precision
If practitioners review thousands of forecast models manually, then comprehensive analysis may be achieved, but time consumption increases significantly
Solution Approach 1:
The patent performs preliminary analysis to identify forecastable patterns before comprehensive model creation and review. By pre-identifying patterns with desirable variability characteristics, the system reduces the number of models that require manual review, saving time while maintaining analysis completeness through automated pattern recognition.
3Loss of information
If all time series data is analyzed without filtering, then no useful data is missed, but evaluation resources are wasted on unforecastable series
Solution Approach 1:
The patent performs preliminary filtering of time series data based on variability measures and pattern recognition before comprehensive evaluation. This preliminary action identifies and eliminates unforecastable series early in the process, preventing waste of evaluation resources while maintaining data completeness for forecastable patterns.
Solution Approach 2:
The patent extracts and removes unforecastable time series patterns from the analysis set based on variability thresholds and pattern matching. By taking out these unforecastable series before detailed evaluation, the system improves evaluation efficiency while maintaining completeness for the remaining forecastable patterns.
Data Source
AI summary
A method, computer program product, and/or computer system improves a future efficiency of a specific system. One or more processors receive multiple historical data snapshots that describe past operational states of a specific system. The processor(s) identify a time series pattern for the time series of data in the multiple historical snapshots and calculate their variability. The processor(s) then determine that the variability in a first sub-set of the time series pattern is larger than a predefined value, and determine that future values of the first set of the time series pattern are a set of non-forecastable future values. The processor(s) also determine that the variability in a second sub-set of the time series pattern for the data is smaller than the predefined value, and utilizes this second sub-set to modify the specific system at a current time.