AI-Generated Simulation Code for Industrial Control Logic Testing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Creating a realistic simulation environment for testing control logic code in automated industrial processes is complex, time-consuming, and requires substantial domain expertise, especially when chemical reactions or flow models are involved, making it costly and inefficient.
Innovation Solution
A method using generative AI models, specifically large language models (LLMs), to generate simulation code for testing control logic code by processing project context embeddings and requirement artifacts, enabling rapid and flexible simulation environment creation with limited specialized expertise.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a realistic simulation environment is created for testing control logic code, then the functional correctness and safety of control logic can be ensured, but the process becomes complex, time-consuming, and requires substantial domain expertise
Solution Approach 1:
The patent uses generative AI to create simplified copies or representations of the industrial process rather than building complex realistic simulations. The AI generates simulation code that captures essential process behavior without requiring detailed domain expertise, thus ensuring functional correctness while reducing complexity
Solution Approach 2:
The patent replaces the traditional manual mechanical process of simulation environment creation (requiring domain experts to manually configure complex simulations) with an AI-based system that automatically generates simulation code from process descriptions, eliminating the need for substantial human domain expertise
2Reliability
If a realistic simulation environment is created for testing control logic code, then the functional correctness and safety of control logic can be ensured, but the process becomes time-consuming and costly
Solution Approach 1:
The AI system generates simulation code rapidly by creating simplified representations of the process, avoiding the time-consuming manual configuration of detailed realistic simulations while still providing sufficient fidelity for functional testing
Solution Approach 2:
The generative AI system performs the simulation environment creation task autonomously without requiring time-intensive human expert intervention. The AI self-generates the simulation code from process descriptions, dramatically reducing the time required to set up testing environments
3Reliability
If simulation libraries for chemical processes are created with sophisticated procedures for emulating chemical reactions, then realistic process simulation is achieved, but the process requires specialists rather than automation providers
Solution Approach 1:
The patent replaces the complex manual process of creating sophisticated simulation libraries with AI-based automatic generation. The AI system handles the complexity of emulating chemical reactions and process behavior, allowing automation providers without deep domain expertise to create accurate simulations
Solution Approach 2:
The AI generates simulation code that captures essential process behavior through simplified representations, achieving sufficient realism for testing purposes without requiring the creation of highly detailed sophisticated simulation libraries by specialists
4Reliability
If traditional methods are used to create simulation environments, then domain expertise ensures accuracy, but the process is expensive and inefficient
Solution Approach 1:
The generative AI system autonomously generates accurate simulation code from process descriptions without requiring time-intensive human expert review and configuration, dramatically improving the efficiency of simulation creation while maintaining accuracy through AI-based process understanding
Data Source
Figure 1
AI summary
A method for generating a simulation code (190) to test a control logic code (160) is described , wherein the control logic code (160) is configured for controlling an automated industrial process, the method comprising: providing the control logic code (160), wherein the control logic code (160) comprises identifiers, which are related to a plurality of components of the automated industrial process; generating a simulation generation prompt (170) for a generative artificial intelligence model configured for natural language processing (150), based on the control logic code (160); performing a similarity search by means of a database (122) and based on the simulation generation prompt (170), wherein the database (122) comprises embeddings of a plurality of elements of a project context (126) of the automatic industrial process; generating an augmented simulation generation prompt (180), wherein the simulation generation prompt (170) is augmented based on results of the similarity search; providing the augmented simulation generation prompt (180) to the generative artificial intelligence model (150); and generating the simulation code (190) for testing the control logic code (160) by means of the generative artificial intelligence model (150), based on the augmented simulation generation prompt (190).