AI-Assisted Industrial Software Generation from Process Mining
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The aging workforce in process automation and control engineering is leading to a loss of knowledge and experience, resulting in prolonged durations and increased errors in designing software applications for industrial processes due to the complexity and variety of factors involved.
Innovation Solution
A method utilizing artificial intelligence (AI) and process mining to generate assistance data based on previous generations of software applications, enabling users with less experience to efficiently create software applications by leveraging knowledge and experience from retiring engineers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual input from experienced engineers is used to generate software applications, then the quality and reliability of the software are improved, but the time consumption and complexity of the generation process increase
Solution Approach 1:
The system performs preliminary action by automatically generating software application code, configuration files, and documentation before manual review, based on process data and specifications. This pre-generation reduces the time engineers need to spend on routine tasks while maintaining quality through automated best practices.
Solution Approach 2:
An intermediary system (the software generation system) is introduced between the engineer's requirements and the final software product. This intermediary automatically translates process specifications into software code, reducing direct manual effort while preserving engineer oversight for critical decisions.
2Reliability
If manual input from experienced engineers is used to generate software applications, then the accuracy and error reduction are improved, but the device complexity and operational difficulty increase
Solution Approach 1:
The system enables self-service by allowing process data and specifications to automatically generate software applications without requiring engineers to manually write code or configure complex parameters. The system serves itself by using the input data to produce the software artifact with minimal human intervention.
Solution Approach 2:
The system uses copying by replicating proven software patterns, configurations, and best practices from existing reliable software applications. This allows the generation of new software by adapting proven templates, reducing errors while simplifying the generation process through pattern reuse.
3Productivity
If process mining with AI is used to generate assistance data, then the productivity and efficiency of software generation are improved, but the loss of information from retiring engineers occurs
Solution Approach 1:
The system implements feedback by continuously learning from generated software and engineering corrections, using this feedback to improve future generations. This creates a knowledge accumulation loop where experience is captured and refined over time, preventing knowledge loss while improving productivity.
Solution Approach 2:
The system applies parameter changes by transforming unstructured engineer knowledge into structured data parameters that can be processed by AI models. This conversion preserves knowledge in a machine-readable format, enabling automated generation while retaining the essence of expert knowledge.
Data Source
AI summary
A method for assisting a user in computer-aided generation of at least a part of a software application for an industrial process includes obtaining assistance data for assisting the user in the computer-aided generation of the at least part of the software application, wherein the assistance data is based on process mining of computer-aided generation data, the computer-aided generation data relating to previous computer-aided generations of at least parts of software applications for other industrial processes by one or more users, the process mining being at least supported by an artificial intelligence (AI) model; and providing the assistance data for assisting the user in computer-aided generation of the at least part of the software application by the user.