Generative AI and PLC coding with automation beyond the basics

Siemens unveiled the first generative AI product tailored for engineering in an industrial environment, called the Siemens Industrial Copilot at Hannover Messe 2024. Integrated with Siemens TIA Portal, Industrial Copilot is changing how engineering teams approach PLCs. Here, Tom Cash, director of automation parts supplier Foxmere, explains how the shift from theory to practical application of generative AI-powered tools will benefit engineers.

The most common programming language for PLCs is ladder logic (LL), which mimics the appearance of electrical relay logic diagrams. Engineers manually design the logic by arranging symbols, such as contacts, coils, timers and counters, in a graphical format representing the sequence of operations in a control process.

With the integration of generative AI, the conventional process of PLC programming is evolving, presenting engineers with powerful new tools to streamline and enhance their work. As Dr Steven Althaus of Grenzebach Group implies, generative AI isn't just a tool; it's a "must-have."

Thanks to the Industrial Copilot software, engineers can quickly locate the right help topics, generate basic visualisations and expedite PLC code development. This marks a notable leap forward in the application of AI in industrial automation, but what are the benefits?

To clarify, Siemens Industrial Copilot uses advanced AI algorithms to generate PLC code in Structured Control Language (SCL). This is a key development because it eradicates the necessity for engineers to manually write and copy-paste code. The software offers step-by-step guidance on repairs and will even suggest ideas to improve broader system performance.

The software also suggests code snippets within the TIA Portal, simplifying the coding process and reducing the probability of human error. As mentioned, the AI can also create visualisations for machines or plants utilising WinCC Unified, streamlining the creation of intuitive interfaces for monitoring and control.

For instance, in a food processing plant, operators rely on HMIs to monitor production lines, track ingredient usage and ensure compliance with safety regulations. Customising these HMIs to suit the specific needs of different products or production lines can be challenging. The software can assist engineers by generating initial HMI designs based on a line's requirements. These designs can then be refined to meet specific operational needs.

Beyond code generation, Siemens Industrial Copilot presents valuable assistance by explaining SCL code blocks and guiding engineers through complex tasks. This capability helps demystify intricate code and supports engineers in understanding and implementing complex automation logic. The app also features a natural language search for Siemens manuals, allowing users to quickly find relevant information without going through extensive documentation. This is particularly important during a time where industries are facing a shortage of skilled labour because this software can act as a digital assistant, guiding less experienced engineers and operators through complicated tasks.

Siemens Industrial Copilot is integrated with Siemens Xcelerator, a digital business platform with automation and process simulation information. This allows the AI to analyse data from across the factory and suggest optimisations, making industrial processes more efficient. By automating repetitive tasks and providing intelligent code suggestions, Copilot significantly reduces the engineering workload. This saves time and makes the process less prone to human error, leading to more reliable automation solutions. Put simply, the result is shorter development times, a codebase with fewer bugs and increased productivity. For example, generative AI could analyse vast amounts of production data to uncover patterns in a wind turbine manufacturing facility. Here, it might recommend adjusting the speed of a particular machine to balance energy efficiency with output, leading to significant cost savings and a more sustainable operation.

Most of today's PLCs execute pre-programmed instructions, often in LL or other structured languages. Usually used in conjunction with SCADA systems, PLCs monitor inputs from sensors, make logical decisions, and control outputs to actuators, like motors. That's why selecting the right controller for your application is considered the fundamental step in ensuring optimal performance and reliability in your system. Historically, PLCs revolutionised industrial automation by replacing cumbersome relay-based systems with programmable controllers capable of executing complex control sequences.

The integration of generative AI technology in the Siemens Industrial Copilot marks a new chapter in this evolution, which will benefit engineers and manufacturers alike.