New approaches in Machine Automation: Why process models create sustainable value
Optimizing machines and production processes is a crucial factor for a company's success across the entire industry. In this blog post, we explain...
3 min read
Sabine Reisinger : May 8, 2024 8:54:29 AM
In this interview, you'll learn more about how our technology differs from AI in PLC programming and how it overcomes the limitations of artificial intelligence. We'll also discuss whether AI could replace humans in machine programming.
Sabine Reisinger: Hello, Markus. Today, we want to talk about Selmo and artificial intelligence in relation to PLC programming. Why is AI, particularly in this field, such a polarizing topic nowadays?
Markus Gruber: In recent years, there have been tremendous advancements in the development of artificial intelligence, extending even into personal applications. I'm sure each of us has tried entering something into ChatGPT, right?
Artificial intelligence is also a very current topic in machine programming. AI is capable of solving tasks faster than traditional programming methods. For example, you can use ChatGPT to automatically generate control codes. However, whether these codes are correct is another matter, as AI also has its limitations.
Sabine Reisinger: I would like to build on the last point. Would you say that AI also has limitations in PLC programming?
Markus Gruber: Certainly, AI also has its limitations. You've probably entered a question into ChatGPT and received an answer before. If you were to ask the same question again, there's a high likelihood that you would not receive the exact same answer.
AI systems learn from historical data and methods. The effectiveness of AI depends on the quality of the available data. This is especially true in PLC programming. AI systems are trained to recognize patterns and make predictions based on historical data. Based on this data, they provide only an interpretation of how a machine might function.
We must also not forget the uncertainties regarding data protection and legal frameworks, which are crucial when dealing with sensitive data such as machine programming.
Sabine Reisinger: What could be the consequences of relying solely on artificial intelligence for PLC code generation?
Markus Gruber: One obvious consequence would be coding errors, which in turn would require a lot of time for monitoring and control to detect these errors. There's also the possibility that the code might not work at all, and thus the machine wouldn't either. Programs are not reliably testable, functions cannot be fully verified, and there may be difficulties in maintaining future machines. Moreover, each change results in a new program generated by AI.
Ultimately, the use of AI leads to human expertise being pushed into the background.
Sabine Reisinger: You developed Selmo Technology several years ago, which has revolutionized manual programming because no manual coding is required anymore. How do you differentiate this from the use of AI?
Markus Gruber: Selmo bridges the gap between humans and machines. With our technology, the process of a machine is modeled, after which Selmo automatically generates precise and error-free PLC code. Since we rely on a unique algorithm, we are not dependent on extensive datasets or learning curves like AI. This means that when you model the process in the Selmo Studio, you always get the same result.
Sabine Reisinger: What advantages do companies gain by using Selmo for PLC programming of their machines?
Markus Gruber: Short and to the point: Ready-to-use, error-free PLC code and functioning, reliable machines!
Sabine Reisinger: How do you see the future of Selmo and artificial intelligence in machine automation?
Markus Gruber: With Selmo, we focus on behavior-oriented engineering! This is a very important factor because for future digitalization and the development of AI models, not only functional data but also behavioral data are crucial. Our technology lays the foundation for AI analysis in the OT area and simultaneously creates a uniform interface for IT. This means that our solution enables digitalization and the development of AI models in the OT area, making every digital information from this sector automatically available. Thus, Selmo provides the basis for future AI models for process optimization.
Sabine Reisinger: A final and I think very frequently asked question: Do you believe that AI will one day replace humans in machine programming?
Markus Gruber: You're right... I've heard that question so many times (laughs). My answer is no. I believe that AI will not completely replace humans in machine programming. In mechanical engineering, the true creative power lies with humans. Expertise, contextual understanding, and creativity will always be important in developing innovative solutions. Relying solely on past data is not enough!
**Selmo is an OT (Operational Technology) platform for real-time deviation detection. Our technology generates deterministic programs based on models and algorithms and detects unknown errors as they occur, all without any prior programming of the error message.
Optimizing machines and production processes is a crucial factor for a company's success across the entire industry. In this blog post, we explain...
The persistent Challenge for companies Manufacturing companies significantly contribute to the competitiveness of a country. However, they are...
Standardization in machine programming In mechanical engineering, there is no generally applicable industry standard, especially in the programming...