Predictive Fault Modeling for Process Deviation Detection

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing methods for predicting process deviations in process engineering plants, such as painting plants, are not reliable and efficient, lacking a systematic approach for early detection and analysis of fault causes.

Innovation Solution

A method for predicting process deviations using a predictive model based on machine learning, which includes automatic detection of fault situations, determination of fault causes, and prioritization of process values, utilizing a fault database and historical data for anomaly detection and fault analysis.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If traditional error prediction methods are used, then the system can detect faults, but the prediction reliability and efficiency are insufficient

Engineering Contradiction:
Improveprediction reliabilityVSAvoidprediction efficiency
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The system performs preliminary actions by storing historical fault data and training the neural network model in advance. The model is pre-trained with balanced data containing both normal and fault conditions, enabling early and reliable fault prediction before actual deviations occur, thus improving both reliability and efficiency.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent replaces traditional mechanical or rule-based fault detection systems with an intelligent neural network model. This substitution enables the system to automatically learn patterns from historical data and make accurate predictions, significantly improving prediction reliability while maintaining high efficiency through automated processing.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

2Measurement precision

If more process values are monitored to improve detection accuracy, then the measurement precision increases, but the system complexity increases

Engineering Contradiction:
Improvefault detection precisionVSAvoidsystem complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The neural network model serves as a universal processing unit that handles multiple process values simultaneously. Instead of requiring separate analysis methods for each parameter, the single model processes temperature, humidity, pressure, and other variables together, improving detection precision without proportionally increasing system complexity.

Inventive Principle:
Principle #6Universality (Multi-functionality)

Solution Approach 2:

The patent merges multiple data processing functions into a unified neural network framework. Historical data storage, feature extraction, pattern recognition, and prediction functions are combined in one integrated system, allowing comprehensive monitoring of multiple process values while maintaining manageable system complexity through unified architecture.

Inventive Principle:
Principle #5Merging (Combining)

Data Source

PatentEP3966649B1Prediction method and devices for same
Publication Date: 2026.03.25 DUERR SYST AG
  • EP3966649B1 patent drawingFigure 1
  • EP3966649B1 patent drawingFigure 2
  • EP3966649B1 patent drawingFigure 3

AI summary

The invention relates to a method for predicting process deviations in a processing system, for example in a painting system, by means of which process deviations can be easily and reliably predicted, said method comprising the following: - automatically creating a prediction model; - predicting process deviations during operation of the processing system using the prediction model.