Fault-Augmented Manufacturing Models for Faster Root Cause Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current root cause analysis (RCA) methods for manufacturing systems face challenges due to the complexity of cyber physical manufacturing systems, requiring extensive labeled data for model-based approaches and being constrained by limited data availability, making it difficult to manually construct system models and perform fault diagnosis efficiently.
Innovation Solution
A hybrid data-driven and model-based approach is employed to construct a fault-augmented system model, using parameterized fault models to simulate faults and detect changes in production line topology, enabling efficient fault diagnosis and root cause analysis by generating causality graphs and adding causality links through fault scenario simulations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If model-based approaches are used for root cause analysis, then knowledge about system and physics of failure can be utilized, but extensive labeled data is required which is often unavailable
Solution Approach 1:
The patent introduces an intermediary component - a system model constructed from operational data that mediates between available operational data and the requirements of model-based fault analysis. This system model serves as a bridge, allowing fault scenario simulations to be performed without requiring extensive labeled fault data, thus resolving the contradiction between utilizing physics-based knowledge and data availability constraints
Solution Approach 2:
The patent applies preliminary action by constructing the system model and performing fault scenario simulations in advance. The system model is built from operational data before actual fault analysis is needed, and fault scenarios are pre-simulated to establish causality relationships. This preliminary preparation enables rapid root cause analysis when faults occur without requiring extensive labeled fault data at the time of analysis
2Measurement precision
If physics-based models are constructed with high detail, then accurate fault analysis can be performed, but the complexity and cost of model construction becomes prohibitive
Solution Approach 1:
The patent applies partial action by constructing a system model that includes only the necessary components and relationships relevant to fault analysis, rather than modeling every detail of the physical system. The model captures essential operational relationships and fault propagation paths without requiring complete physical fidelity, thus achieving sufficient accuracy while reducing complexity and construction cost
Solution Approach 2:
The patent segments the system model into modular components representing different operational elements and their relationships. This segmentation allows the model to be constructed from available operational data in a structured manner, reducing overall complexity while maintaining the ability to perform accurate fault analysis through composition of these modular segments
3Adaptability or versatility
If manual construction of system models is performed, then customization to specific systems is possible, but the process becomes infeasible for complex cyber physical manufacturing systems
Solution Approach 1:
The patent enables the system model to be constructed automatically from operational data without requiring manual intervention. The system serves itself by extracting relationships and parameters directly from operational data, performing fault scenario simulations autonomously, and generating the system model structure automatically. This self-service approach maintains adaptability to specific manufacturing systems while eliminating the prohibitive manual construction effort for complex systems
Data Source
Figure 1A
Figure 1B
Figure 1C
AI summary
A system is provided for determining causes of faults in a manufacturing system. The system stores data associated with a processing system which includes machines and associated processes, wherein the data includes timestamp information, machine status information, and product-batch information. The system determines, based on the data, a topology of the processing system, wherein the topology indicates flows of outputs between the machines as part of the processes. The system determines information of machine faults in association with the topology. The system generates, based on the machine-fault information, one or more fault parameters which indicates frequency and severity of a respective fault. The system constructs, based on the topology and the machine-fault information, a system model which includes the one or more fault parameters, thereby facilitating diagnosis of the processing system.