Autonomous Deviation Analysis Using Simulated State Data

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current root cause analysis methods in automation technology and robotics are time-consuming, error-prone, and require expert knowledge, making them inefficient and costly for identifying deviations in systems like industrial robots.

Innovation Solution

An autonomous unit analyzes current and past system data using similarity metrics and simulations to independently determine the cause of deviations, allowing for autonomous error cause analysis and potential countermeasures, reducing reliance on human experts and improving reliability.

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 the cause of deviations, but the analysis becomes time-consuming and complex

Engineering Contradiction:
Improveaccuracy of root cause identificationVSAvoidanalysis time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The autonomous unit performs root cause analysis independently using its own processing capabilities, comparing current state data with historical data and simulations without requiring external expert intervention. The system serves itself by automatically identifying deviations, determining their causes, and implementing countermeasures

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

Manual expert analysis is replaced with an automated information processing system that uses algorithms to compare state data, evaluate similarities, and determine causes. The mechanical process of expert review is substituted with computational methods including data comparison, similarity metrics, and simulation techniques

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

2Reliability

If manual expert analysis is used for root cause identification, then accurate causes can be determined, but expert knowledge is required and costs increase

Engineering Contradiction:
Improveaccuracy of deviation analysisVSAvoiddependency on expert knowledge
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The autonomous unit independently performs the analytical function that previously required external experts. It autonomously compares state data, identifies deviations, determines causes through simulation and comparison, and implements countermeasures without human intervention

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The system creates virtual copies of system states through simulations that replicate actual operating conditions. By comparing current state data with simulated scenarios, the system identifies causes without requiring physical presence or interpretation by experts

Inventive Principle:
Principle #26Copying

3Loss of information

If conventional logging and remote monitoring are used to record system states, then data is available for analysis, but the approaches are insufficient for increasing automation and Industry 4.0 requirements

Engineering Contradiction:
Improveavailability of system dataVSAvoidautonomous analysis capability
Core Design Contradiction:
Loss of informationVSExtent of automation

Solution Approach 1:

The autonomous unit automatically processes recorded data through sophisticated comparison algorithms and simulations, transforming passive data storage into active autonomous analysis. The system self-determines causes and implements countermeasures based on the recorded state data

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The system combines multiple data sources (state data, sensor data, simulation data) and multiple analytical methods (comparison algorithms, similarity metrics, simulations) into an integrated autonomous analysis capability that exceeds the sum of its individual components

Inventive Principle:
Principle #40Composite materials

Data Source

PatentEP3582050B1Method for analysing a cause of at least one deviation
Publication Date: 2021.04.28 SIEMENS AG
  • EP3582050B1 patent drawingFigure 1
  • EP3582050B1 patent drawingFigure 2

AI summary

The invention relates to a method for analyzing a cause of at least one deviation, comprising the steps of: receiving a state data set (10) to be analyzed, comprising the at least one deviation (S1); determining at least one previous state data set (20, S2); determining at least one alternative previous state data set (30) based on the at least one previous state data set (20, S3); determining at least one simulated data set (40) by simulating the at least one alternative previous state data set (30, S4); comparing the at least one simulated data set (40) with the state data set to be analyzed (10, S5); determining a similarity value between the at least one simulated data set (40) and the state data set to be analyzed (10, S6);Output of at least one simulated data set (40), at least one alternative preceding state data set (30) as the cause of at least one deviation or at least one error message depending on the similarity value (S7). The invention further relates to a corresponding autonomous unit and a corresponding computer program product.