Autonomous Deviation Analysis Using Simulated State Records

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current root cause analysis methods in automation and robotics are time-consuming, error-prone, and require expert knowledge, making them inadequate for increasing automation and Industry 4.0 applications.

Innovation Solution

An autonomous unit analyzes deviations by receiving state data records, determining preceding records, simulating alternative scenarios, comparing them for similarity, and outputting the cause or error messages, allowing for independent and efficient root cause analysis.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If manual root cause analysis is performed using conventional approaches (logging and remote monitoring), then experts can identify causes of deviations, but the process becomes time-consuming and requires expert knowledge

Engineering Contradiction:

Improveaccuracy of root cause identificationVSAvoidtime required for analysis

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The system performs self-diagnosis by automatically analyzing sensor data and simulation results to identify root causes of deviations, eliminating the need for external expert intervention and significantly reducing analysis time while maintaining high accuracy

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The system pre-calculates and stores multiple simulated state data records representing different possible failure scenarios before actual deviations occur. When a deviation is detected, the system quickly compares actual data against these pre-prepared simulations to rapidly identify the root cause

Inventive Principle:

Principle #10Preliminary action

2Reliability

If manual root cause analysis is performed by experts, then accurate cause identification is possible, but the process becomes complex and error-prone

Engineering Contradiction:

Improveaccuracy of root cause identificationVSAvoidcomplexity of analysis process

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The system replaces the manual mechanical process of expert analysis with an automated computational system that uses algorithms to compare sensor data against simulated scenarios, eliminating human error and simplifying the complex analysis process into systematic automated comparisons

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system creates virtual copies of system states through simulation, generating simulated state data records that replicate actual system behavior under various conditions. This allows automated comparison and analysis without requiring experts to manually recreate or analyze complex physical scenarios

Inventive Principle:

Principle #26Copying

3Loss of information

If conventional logging and remote monitoring are used to store system states and measurement data, then a basis for root cause analysis is available, but the approaches are inadequate for increasing automation

Engineering Contradiction:

Improvecompleteness of data for analysisVSAvoidlevel of autonomous analysis

Core Design Contradiction:

Loss of informationVSExtent of automation

Solution Approach 1:

Inventive Principle:

Principle #26Copying

Solution Approach 2:

The system continuously monitors sensor data, compares it against simulated scenarios, and automatically adjusts its analysis based on the results. This closed-loop feedback mechanism enables the system to autonomously refine its root cause identification without human intervention

Inventive Principle:

Principle #23Feedback

Data Source

PatentUS11138057B2Method for analyzing a cause of at least one deviation

Publication Date: 2021.10.05 SIEMENS AG

AI summary

Provided is a method and corresponding unit for analyzing a cause of at least one deviation, having the steps of: receiving a state data record which has at least one deviation; determining at least one preceding state data record; determining at least one alternative preceding state data record based on the at least one preceding state data record; determining at least one simulated data record by simulating the at least one alternative preceding state data record; comparing the at least one simulated data record with the state data record to be analyzed; determining a similarity value between the at least one simulated data record and the state data record; outputting the at least one simulated data record of the at least one alternative preceding state data record as the cause of the at least one deviation or at least one error message on the basis of the similarity value.