Industrial Machine Event Prediction with Sensor Channel Failover
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Identifying actionable data points for industrial machines under varying operating conditions is challenging, especially when sensor signaling channels fail or transmit inaccurate data, leading to inaccurate insights and increased maintenance complexity.
Innovation Solution
Implementing machine-learning (ML)-assisted failover and relabeling of sensor signaling channels using neural networks to generate predicted values 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
1Reliability
If sensor signaling channels are used to monitor industrial machines, then condition monitoring capability is improved, but false positives and negatives increase when channels fail or transmit inaccurate data
Solution Approach 1:
A neural network model is introduced as an intermediary between sensor channels and event identification. The model receives data from multiple sensor channels, processes it through learned patterns, and produces event predictions. This intermediary layer filters out noise and compensates for individual channel failures, reducing false positives and negatives while maintaining monitoring capability.
Solution Approach 2:
The system implements feedback through continuous monitoring of sensor channel quality and performance. When channel failures or inaccuracies are detected, the system adjusts its processing accordingly, using the neural network to compensate for degraded channels and maintain accurate event identification despite individual channel issues.
2Reliability
If multiple sensor signaling channels are deployed to improve monitoring coverage, then condition monitoring capability is improved, but system complexity increases
Solution Approach 1:
Multiple sensor channels are merged into a unified neural network processing pipeline. Instead of independently analyzing each channel, the system combines their inputs and processes them together through shared model layers, reducing the complexity of managing individual channels while maintaining comprehensive monitoring coverage through the integrated approach.
3Ease of manufacture
If traditional maintenance approaches are used, then maintenance simplicity is maintained, but maintenance effectiveness decreases due to inability to predict equipment-related events
Solution Approach 1:
The neural network model performs preliminary analysis of sensor data to predict equipment-related events before they occur. By identifying patterns and trends in advance, the system enables proactive maintenance planning, allowing maintenance teams to prepare and respond effectively while maintaining relatively simple maintenance procedures.
Data Source
AI summary
Techniques are disclosed herein for machine-learning (ML)-assisted event prediction for industrial machines. A first set of embeddings can be generated based on labeled first event data, which can be labeled with classifiers determined based on signaling channel information for the first event data. A neural network can be trained, using the classifiers, to generate (i) a similarity score for the first set of embeddings and the second set of embeddings and (ii) a classifier recommendation for the second set of embeddings. The second set of embeddings can be generated based on data collected using condition monitoring sensors for a particular industrial machine. Accordingly, the system can generate alerts, recommendations, and/or notifications based on the automatically classified data encoded in the second set of embeddings. Incremental training techniques are disclosed for further training the neural network to minimize false positives and/or false negatives.