Autoencoder Anomaly Prediction for Manufacturing Subsystems
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Manufacturers face challenges with scheduled and reactive maintenance approaches, leading to lost revenue, excess costs, shortened asset life, poor product quality, and safety risks due to unplanned maintenance in processing pipelines.
Innovation Solution
A system that predicts and detects anomalies in processing pipelines using machine learning methods, specifically by determining indicative tags from sensor data, processing these with autoencoders to identify differences and transmit alerts for potential failures, enabling predictive maintenance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If scheduled and reactive maintenance is used to manage operating assets, then maintenance can be performed, but lost revenue from deferred production, excess costs, shortened asset life, poor product quality, and personnel safety risks occur
Solution Approach 1:
The system performs preliminary action by predicting asset failures before they occur using machine learning models that analyze historical and real-time sensor data. The anomaly detection system identifies patterns indicating potential failures, allowing maintenance to be scheduled proactively rather than reactively, thus preventing production interruptions and safety incidents while optimizing maintenance timing to avoid unnecessary downtime
2Measurement precision
If more sensors and data collection are deployed to improve anomaly detection accuracy, then prediction accuracy improves, but system complexity and data processing requirements increase
Solution Approach 1:
The system extracts only the most relevant features and tags from the vast sensor data using machine learning techniques. The autoencoder model automatically identifies and extracts key patterns from hundreds or thousands of sensor inputs, filtering out redundant information. This extraction approach maintains high anomaly detection accuracy while reducing the effective data dimensionality and computational complexity required for processing
Solution Approach 2:
The autoencoder serves as an intermediary between raw sensor data and anomaly detection. It compresses high-dimensional sensor data into a lower-dimensional latent representation, then reconstructs it to identify anomalies. This intermediary layer simplifies the data processing pipeline by transforming complex multi-sensor inputs into manageable features that feed into the anomaly detection logic, reducing overall system complexity
Data Source
AI summary
The present disclosure provides system, methods, and computer program products for predicting and detecting anomalies in a subsystem of a system. An example method may comprise (a) determining a first plurality of tags that are indicative of an operational performance of the subsystem. The tags can be obtained from (i) a plurality of sensors in the subsystem and (ii) a plurality of sensors in the system that are not in the subsystem. The method may further comprise (b) processing measured values of the first plurality of tags using an autoencoder trained on historical values of the first plurality of tags to generate estimated values of the first plurality of tags; (c) determining whether a difference between the measured values and estimated values meets a threshold; and (d) transmitting an alert that indicates that the subsystem is predicted to experience an anomaly if the difference meets the threshold.