Sensor Anomaly Detection Using Cross-Sensor Regression Forecasting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing predictive maintenance techniques in the field of anomaly detection for component failures are inadequate when historical sensor data is insufficient, as they rely on pattern identification from past failures, which may not be available in sufficient quantities.
Innovation Solution
The system employs regression calculations based on previous sensor values to determine expected values, allowing for efficient forecasting and anomaly detection by comparing current sensor values, thereby improving predictive maintenance even with limited historical data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If historical sensor data is used for pattern identification to detect component failures, then anomaly detection accuracy is improved, but the method becomes unsuitable when sufficient historical data is not available
Solution Approach 1:
The patent transforms the approach from pattern matching (which requires abundant historical data) to regression-based forecasting (which can work with limited data). By changing the fundamental parameter of the detection method from statistical pattern recognition to mathematical regression modeling, the system can generate reliable predictions even when historical data quantities are insufficient for traditional anomaly detection techniques.
2Reliability
If regression calculations are used for forecasting with limited historical data, then predictive maintenance capability is improved, but the method requires efficient computation to handle real-time sensor data
Solution Approach 1:
The patent extracts and applies regression calculations specifically for generating expected sensor values, separating this forecasting function from the broader anomaly detection system. By isolating the regression computation as a distinct module that takes sensor data as input and produces expected values for comparison, the system achieves computational efficiency while maintaining reliable predictive maintenance capability.
Data Source
AI summary
Some embodiments include reception of a time-series of a respective data value generated by each of a plurality of sensors, calculation of a regression associated with a first sensor of the plurality of sensors based on the received plurality of time-series, the regression being a function of the respective data values of the others of the plurality of data sources, reception of respective data values associated with a time from and generated by each the plurality of respective sensors, determination of a predicted value associated with the time for the first sensor based on the regression associated with the first sensor and on the respective data values associated with the time, comparison of the predicted value with the received value associated with the time and generated by the first sensor, and determination of a value indicating a likelihood of an anomaly based on the comparison.