AI Planning for Heterogeneous Automation Configuration Data

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Different automation systems have varying structures, making it difficult to compare and utilize their data for planning and setup, as existing assistance systems lack effective methods to handle the heterogeneity and complexity of configuration data.

Innovation Solution

An assistance system utilizing artificial intelligence and machine learning with a neural network architecture for deep learning processes configuration data from diverse automation systems, normalizing and analyzing it to support planning and setup by transforming data into a uniform format for better comparison and analysis.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If configuration data from different automation systems is directly compared and analyzed, then the planning and setup process can be supported, but the heterogeneity and complexity of the data structures make direct comparison difficult

Engineering Contradiction:
Improveability to process configuration data from different automation systemsVSAvoidcomplexity of processing heterogeneous configuration data
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent transforms heterogeneous configuration data into a unified vector representation, changing the parameter format from diverse structured data to standardized numerical vectors. This allows the AI component to process configuration data from different automation systems uniformly, resolving the contradiction between adaptability and complexity.

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent introduces an AI component with a neural network as an intermediary between the heterogeneous configuration data and the analysis process. This intermediary automatically learns and adapts to different data structures, enabling the system to handle diverse automation system configurations without requiring complex manual processing rules.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Reliability

If more configuration data from diverse automation systems is processed to improve planning support, then the quality of recommendations improves, but the complexity and difficulty of processing the data increases

Engineering Contradiction:
Improvequality of planning supportVSAvoiddifficulty of processing complex configuration data
Core Design Contradiction:
ReliabilityVSDifficulty of detecting and measuring

Solution Approach 1:

The patent replaces traditional mechanical data processing methods with an AI-based neural network system. Instead of using predefined rules and manual processing to handle diverse configuration data, the system uses machine learning to automatically detect patterns and extract meaningful information, reducing the difficulty of processing while improving reliability.

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

Solution Approach 2:

The patent transforms complex configuration data into simplified vector representations that capture essential features while discarding redundant information. This parameter transformation makes the data more suitable for AI processing and enables the system to handle diverse configurations with consistent methods, improving both reliability and reducing processing difficulty.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP3575890B1Assistance system for supporting planning of automation systems
Publication Date: 2022.03.30 SIEMENS AG
  • EP3575890B1 patent drawingFigure 1
  • EP3575890B1 patent drawingFigure 2~3
  • EP3575890B1 patent drawingFigure 4

AI summary

The present invention describes an assistance system (100) for supporting the planning of automation systems, comprising: - a configuration database (600) comprising configuration data records (610) of automation systems, wherein each configuration data record (610) comprises the configuration data of a given automation system; - a vectorization component (300) for structuring and adapting configuration data records, wherein the vectorization component (300) is designed and configured to convert the configuration data records (610) of the configuration database (600) into vectorized configuration data records (310); and wherein the assistance system (100) further comprises an AI component (400) for processing the vectorized configuration data records (310) using artificial intelligence.wherein the processing of the vectorized configuration datasets (310) by the AI ​​component (400) involves the use of a neural network (500) with a deep learning architecture.