Automation Assembly RCA via Recursive Neighbor Interrogation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing industrial automation systems face challenges in performing root cause analysis (RCA) in environments with dynamic structures or frequent reconfigurations, particularly when plants have not been fully modeled or when production shifts between different products.
Innovation Solution
A method and system that enable decentralized root cause analysis by allowing automation components to dynamically discover their plant structure and perform local calculations, reducing the need for manual modeling and external coordination.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a plant is manually modeled in software tools with all machines and connections created, then root cause analysis can be performed on external computers, but the process requires very considerable manual effort and time, and data are often not specified by machine manufacturers
Solution Approach 1:
The automation components perform self-identification and self-description, automatically providing their identity information and behavior models to neighboring components without requiring manual modeling. Each component autonomously generates and transmits its own data, eliminating the need for external manual plant modeling while maintaining high accuracy for root cause analysis.
2Loss of information
If all relevant key figures of all machines are stored and continuously reported to a SCADA system, then real-time data are available for analysis, but the data infrastructure becomes complex and requires continuous configuration updates when plant structure changes
Solution Approach 1:
The system divides the data collection and processing function into autonomous segments at each automation component level. Each component independently manages its own data generation, storage, and transmission to neighbors, eliminating the need for a centralized complex data infrastructure while ensuring continuous availability of dynamic data for real-time analysis.
3Adaptability or versatility
If the plant structure is frequently reconfigured for different products, then production flexibility is improved, but the manual modeling must be updated continuously which is very time-consuming
Solution Approach 1:
The system implements dynamic self-configuration where automation components automatically detect structural changes in the plant and update their connectivity information in real-time. When the plant structure changes due to reconfiguration for different products, components autonomously re-establish their neighbor relationships and data exchange paths without requiring manual model updates, thus maintaining production flexibility without time loss.
4Ease of operation
If RCA is performed on an external computer with centralized data collection, then analysis can be conducted, but the system requires continuous configuration and updates when automation components are added or removed
Solution Approach 1:
Each automation component autonomously maintains its own configuration data and automatically updates its connectivity information when structural changes occur. The decentralized architecture eliminates the need for centralized configuration management, allowing RCA to be executed simply by any component using locally available data from its neighbors without requiring continuous external configuration updates.
Data Source
AI summary
A method and an automation component for identifying a process-disrupting automation component in an industrial automation assembly, wherein a process disruption is determined in a first automation component and examined by a first local analysis device, where an automation component arranged upstream and/or an automation component arranged downstream is first determined by each automation component, an interrogation message is sent from a first automation component to a second automation component and the same interrogation message or a further interrogation message is recursively sent by the second automation component to a third automation component arranged upstream or downstream of the second automation component and processed, where in the event of a locally determined disruption, the relevant automation component sends a response message, which is back-propagated to the origin and signaled such that a decentralized error analysis becomes possible, even with a changing system topology, without the need for any redesign work.

