ML Sensor Validation for Fault Remediation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing fault-remediation systems in monitored systems often fail to accurately address actual faults due to anomalies caused by faulty sensors, potentially exacerbating system failures, and require human intervention to verify sensor data, which can lead to lapses in monitoring and remediation effectiveness.
Innovation Solution
A machine learning model is trained on historical sensor data to validate and estimate sensor values, generating a filtered data set that identifies anomalies and creates task lists for remediation, including both computer-executed and human-executed tasks, to address faults in monitored systems.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If human supervision is used to verify sensor data and perform remediation actions, then reliability of fault remediation is improved, but productivity decreases due to manual intervention requirements
Solution Approach 1:
A machine learning model serves as an intermediary between sensor data and remediation decisions. The model validates sensor readings by comparing them against learned patterns from historical data, estimating true system state when sensors are faulty. This automated intermediary maintains reliability by accurately distinguishing real faults from sensor errors while enabling high-speed automated remediation without human intervention.
Solution Approach 2:
The patent replaces the mechanical human verification process with an automated machine learning-based validation system. Instead of human operators manually checking sensor readings and deciding on remediation actions, the system uses trained models to automatically validate sensor data, detect anomalies, and trigger appropriate remediation workflows, thereby maintaining reliability while dramatically improving productivity.
2Productivity
If automated computer-based monitoring and remediation is implemented, then productivity is improved, but reliability deteriorates when sensors are faulty leading to incorrect remediation actions
Solution Approach 1:
The system performs preliminary validation of sensor data using machine learning models before triggering remediation actions. By pre-training models on historical sensor data and normal system behavior patterns, the system can predict whether a sensor reading represents a real fault or a sensor malfunction. This preliminary check prevents incorrect automated remediation actions while maintaining high productivity.
Solution Approach 2:
The system incorporates feedback mechanisms where the machine learning model continuously learns from validated sensor data and remediation outcomes. The model uses feedback from human-verified cases to improve its ability to distinguish real faults from sensor errors, progressively enhancing reliability while maintaining automated high-speed operation.
3Measurement precision
If sensor data is validated using machine learning models, then measurement precision of fault detection is improved, but device complexity increases due to model training and validation requirements
Solution Approach 1:
The validation system is segmented into modular components: data collection modules, model training modules, validation modules, and remediation modules. Each component handles a specific aspect of the validation process independently. This segmentation allows the system to achieve high measurement precision through specialized processing in each module while managing complexity through modular architecture that can be configured and maintained independently.
Data Source
AI summary
Techniques for using machine learning model validated sensor data to generate recommendations for remediating issues in a monitored system are disclosed. A machine learning model is trained to identify correlations among sensors for a monitored system. Upon receiving current sensor data, the machine learning model identifies a subset of the current sensor data that cannot be validated. The system generates estimated values for the sensor data that cannot be validated based on the learned correlations among the sensor values. The system generates the recommendations for remediating the issues in the monitored system based on validated sensor values and the estimated sensor values.