Visual Automation Code Generation from Industrial Diagrams

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current methods for generating computer-based instruction code require programming expertise, limiting the ability of non-technical individuals, such as machine operators, to create and control automated processes.

Innovation Solution

A system that utilizes digital images of industrial processes to generate automation code, employing artificial intelligence and machine learning to analyze and convert graphical representations into executable code, thereby eliminating the need for programming knowledge.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If traditional programming methods are used to create automation code, then code quality and reliability are improved, but accessibility and ease of operation deteriorate because programming expertise is required

Engineering Contradiction:
Improvecode reliabilityVSAvoidcode creation accessibility
Core Design Contradiction:
ReliabilityVSEase of operation

Solution Approach 1:

The patent introduces an image-to-code conversion system as an intermediary between the operator's visual understanding of the process and the automated code generation. The operator draws a visual representation of the desired automation process, and the system automatically translates this visual input into executable automation code, eliminating the need for the operator to directly write programming code while maintaining code reliability through systematic conversion rules and validation mechanisms

Inventive Principle:
Principle #24Intermediary (Mediator)

2Ease of operation

If automated code generation from images is implemented, then ease of operation and accessibility are improved, but device complexity and system sophistication worsen due to the need for AI/ML components

Engineering Contradiction:
Improvecode creation accessibilityVSAvoidsystem complexity
Core Design Contradiction:
Ease of operationVSDevice complexity

Solution Approach 1:

The patent employs image processing and pattern recognition techniques that analyze the visual drawing created by the operator and generate corresponding automation code through template matching and rule-based conversion. The system creates a digital representation (copy) of the operator's visual intent and systematically transforms it into code, reducing the need for complex AI/ML models while maintaining ease of operation through structured visual-to-code translation

Inventive Principle:
Principle #26Copying

3Manufacturing precision

If manual programming is required for automation tasks, then precision and control over automation logic are improved, but productivity and development time deteriorate

Engineering Contradiction:
Improveautomation logic precisionVSAvoidcode development speed
Core Design Contradiction:
Manufacturing precisionVSProductivity

Solution Approach 1:

The patent enables operators to define automation logic visually through drawings before code generation occurs. This preliminary visual specification allows operators to precisely articulate their automation requirements in a intuitive graphical format, and the system then automatically generates the corresponding precise code, eliminating the time-consuming process of manual programming while maintaining logic precision through the structured visual-to-code translation process

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP4571492A1Generative artificial intelligence for creation of instruction code from an input
Publication Date: 2025.06.18 ROCKWELL AUTOMATION TECH INC
  • EP4571492A1 patent drawingFigure 1
  • EP4571492A1 patent drawingFigure 2
  • EP4571492A1 patent drawingFigure 3

AI summary

Various systems and methods are presented regarding generating executable computer code/instructions from input files, whereby the input files may be an image file (e.g., JPEG, PDF, etc.). The image file can be digital capture of a sequence of instructions such as a graphical representation comprising a ladder diagram, a function block diagram, a sequential function chart, etc. P&ID and suchlike can also be submitted to the system. Code generated from the input files can be enhanced by application of historical data comprising pertinent subroutines, and suchlike. Further, an entity can be prompted to provide further information in the event of the input file does not provide all of the content.