Sensor Channel Failover Using ML-Based Signal Relabeling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems struggle to accurately identify actionable data points for industrial machines under varying operating conditions, particularly when sensor signaling channels fail or transmit inaccurate data, leading to inefficiencies in maintenance and operational monitoring.
Innovation Solution
Implementing machine-learning (ML) assisted failover and relabeling of sensor signaling channels using neural networks to generate predicted values from alternative channels, and incremental training techniques to minimize false positives and negatives, enabling event prediction and condition monitoring for industrial machines.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional sensor monitoring systems are used, then the system structure is simple, but the system cannot accurately identify actionable data points when sensor signaling channels fail or transmit inaccurate data
Solution Approach 1:
A neural network model is introduced as an intermediary between sensor signaling channels and event identification. The model receives input from multiple sensor channels and processes the data to accurately identify events even when individual channels fail or transmit inaccurate data, thereby improving measurement precision without requiring direct sensor-to-event mapping
Solution Approach 2:
The system dynamically adjusts the weights and parameters of the neural network model based on the reliability of different sensor signaling channels. By changing parameters according to channel performance, the system maintains high accuracy in event identification while adapting to varying sensor conditions
2Reliability
If multiple sensor signaling channels are monitored, then the reliability of data collection improves, but the difficulty of detecting and measuring actionable data points increases
Solution Approach 1:
The neural network model segments the analysis of multiple sensor channels by processing each channel's data through dedicated input layers and then integrating the results in hidden layers. This segmentation approach allows the system to handle multiple channels systematically, improving reliability while managing the complexity of detecting actionable data points
Solution Approach 2:
The system implements feedback mechanisms where the neural network continuously monitors the quality and reliability of data from multiple sensor channels, adjusting its processing accordingly. This feedback loop helps distinguish actionable data points from noise even when multiple channels are monitored
3Productivity
If sensor failures are not detected, then the system operates continuously, but false alerts and incorrect maintenance insights occur
Solution Approach 1:
The neural network model is pre-trained with data representing various sensor failure modes and inaccurate data patterns. This preliminary action enables the system to recognize and handle sensor failures proactively, distinguishing between actual equipment issues and sensor problems before false alerts are generated
Solution Approach 2:
The system continuously compares the outputs from multiple sensor channels and uses feedback from the neural network to detect inconsistencies that indicate sensor failures. This real-time feedback mechanism maintains reliable condition monitoring while allowing continuous operation
Data Source
AI summary
Techniques are disclosed herein for machine-learning (ML)-assisted failover and relabeling of sensor signaling channels for industrial machines. A trained neural network can be selectively engaged to generate a set of sensor value predictions for a particular signaling channel using sensor values from another signaling channel (e.g., when the particular signaling channel is down). Using the predicted values, the system can automatically identify and raise alerts regarding machine operating conditions. A first training dataset, used to initially train the neural network, can relate to a set of condition monitoring sensors on a machine and can include a set of input values associatively linked to channel identifiers and/or channel metadata. A second training dataset, generated if a similarity measure between predicted and actual values is under a predetermined threshold, can be used to incrementally retrain the neural network.