Industrial Machine Control via Language Model Command Translation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

The programming and optimization of industrial machines is time-consuming, requiring repeated revisions by engineers to achieve optimal performance.

Innovation Solution

A method utilizing a Language Model, interfaced with a Context Information Library, automatically generates machine commands for industrial machines, leveraging pre-trained AI to understand the machine's working principles and generate optimized sequences of actions.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If traditional programming methods are used by engineers, then the industrial machine can be controlled, but the time required for programming and optimization is excessive

Engineering Contradiction:
Improveprogramming speedVSAvoidtime for programming and optimization
Core Design Contradiction:
ProductivityVSLoss of time

Solution Approach 1:

The patent replaces the mechanical programming process (engineers manually writing and revising code) with an AI-based language model that automatically generates machine commands. The language model interface translates natural language descriptions into executable PLC code, eliminating the need for manual programming and significantly reducing setup time.

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

Solution Approach 2:

The system enables self-service programming where the industrial machine can be programmed and optimized automatically without requiring engineer intervention. The language model generates optimized machine commands autonomously based on task descriptions, allowing the system to program itself and eliminating repetitive manual optimization cycles.

Inventive Principle:
Principle #25Self-service

2Reliability

If engineers repeatedly revise programming to optimize machine performance, then optimal working is achieved, but the process becomes time-consuming

Engineering Contradiction:
Improvemachine performance optimizationVSAvoidtime for repeated revisions
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The language model performs preliminary optimization by generating optimized machine commands in advance, based on the task description and machine context. Instead of requiring multiple revision cycles after deployment, the AI generates near-optimal code initially, reducing the need for repeated revisions and accelerating the time-to-performance.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system incorporates feedback mechanisms where the language model interface receives information about machine performance and operational context, then refines and optimizes the generated commands. This continuous feedback loop enables automatic optimization without manual intervention, maintaining high performance while minimizing revision time.

Inventive Principle:
Principle #23Feedback

3Measurement precision

If custom training of AI models is performed for industrial machine control, then task-specific accuracy is improved, but additional training time and resources are required

Engineering Contradiction:
Improvetask-specific command accuracyVSAvoidtraining time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The patent employs a universal language model that can handle multiple industrial machine control tasks without requiring task-specific training. The model's pre-existing knowledge of programming patterns and its ability to understand natural language descriptions enable it to generate accurate commands for various machine types and applications, eliminating the need for separate training processes for each task.

Inventive Principle:
Principle #6Universality (Multi-functionality)

Solution Approach 2:

The language model interface acts as an intermediary that translates natural language task descriptions into machine-specific commands without requiring direct training on machine data. This intermediary layer adapts the general-purpose language model's capabilities to specific industrial contexts through contextual understanding rather than formal training, saving time and resources.

Inventive Principle:
Principle #24Intermediary (Mediator)

Data Source

PatentEP4592775A1Method of controlling an industrial machine
Publication Date: 2025.07.30 SCHNEIDER ELECTRIC IND SAS
  • EP4592775A1 patent drawingFigure 1
  • EP4592775A1 patent drawingFigure 2
  • EP4592775A1 patent drawingFigure 3

AI summary

The invention relates to a method of controlling an industrial machine, the industrial machine comprising a control unit, e.g. a PLC, and at least one actuator and/or sensor which is controlled by the control unit, wherein - a Language Model is provided, - a Language Model Interface is provided, - a Context Information Library is provided, which stores information on the industrial machine, particularly commands executable by the industrial machine, wherein - the Language Model Interface provides information on the industrial machine from the Context Information Library to the Language Model, - the Language Model sends commands to be executed by the industrial machine to the Language Model Interface, - the Language Model Interface translates the commands received from the Language Model into machine commands for the control unit.