Pneumatic Anomaly Detection With Minimal Sensor Error Localization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for error detection and localization in pneumatic automation plants with minimal sensorics, such as those with only two final position sensors, are inadequate, as they fail to provide reliable and timely defectiveness statements and error localization.
Innovation Solution
An error detection module with a read-in interface for digital signals, a detection algorithm for calculating an anomaly score, and a machine localization method to identify probable causes of errors in automation plant components, utilizing a minimal set of digital signals from final position switches and valve switching commands, and optionally additional sensors for enhanced accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If predictive maintenance algorithms are used with minimal sensor data, then error detection capability is reduced, but system complexity and cost are also reduced
Solution Approach 1:
The system performs preliminary actions by collecting and storing operational data (pressure, flow, temperature, valve positions, actuator positions) during normal operation before failures occur. This historical data is then analyzed using machine learning models to establish baseline behavior patterns, enabling the system to detect anomalies and predict potential failures even with minimal real-time sensor data available.
Solution Approach 2:
The patent introduces an intermediary layer consisting of machine learning models and data processing algorithms that bridge the gap between limited sensor data and comprehensive error detection requirements. These intermediaries synthesize available data, fill information gaps through pattern recognition, and generate meaningful error detection insights without requiring extensive sensor coverage.
2Extent of automation
If machine learning approaches are applied with limited sensor data, then measurement precision deteriorates, but automation extent can be maintained
Solution Approach 1:
The system implements multi-functional data processing capabilities that handle multiple types of sensor data (pressure, flow, temperature, positional information) through a unified machine learning framework. This universal approach allows the system to automatically detect errors across different components and failure modes while maintaining robust performance even with minimal sensor coverage by leveraging patterns across diverse data types.
Solution Approach 2:
The patent dynamically adjusts processing parameters and model configurations based on the specific sensor data available in each application scenario. By changing parameters such as data sampling rates, analysis windows, and model sensitivity thresholds, the system optimizes measurement precision for the given sensor limitations while maintaining high levels of automation.
3Reliability
If redundant monitoring functions are integrated into field devices, then reliability improves, but device complexity increases
Solution Approach 1:
The patent extracts complex monitoring and analysis functions from the field devices themselves and relocates them to a centralized or cloud-based processing system. This extraction allows field devices to remain simple and reliable while the complex redundant monitoring, data analysis, and error detection logic resides in external systems that process data from multiple sources, thereby improving overall reliability without increasing device complexity.
Data Source
AI summary
An error detection and localisation in a pneumatic system and in particular an error detection module includes a read-in interface for reading-in digital signals from the automation plant, a first processor unit designed to execute a detection algorithm for calculating an anomaly score for the automation plant on the basis of the set of read-in signals, a second processor unit which is designed—in the event that the anomaly score calculated with the first processor unit indicates an anomaly—to perform a machine localisation method for localising the error, wherein the machine localisation method has been trained in a training phase in order to calculate and as a result provide, on the basis of a detected circuit diagram of the automation plant with respect to the calculated anomaly score, probabilities of possible causes of error in relation to individual components of the automation plant.