WAGO: Smart Programming Support with AI
Starting Situation
IEC programmers face the daily challenge of using complex, specialized knowledge efficiently. When writing IEC code, they need access not just to extensive technical documentation and specific know-how, but also practical support in the coding process. Relevant knowledge is often spread across various documents, and research takes a lot of time. At the same time, tools that support programmers directly while coding are lacking, such as by providing suitable code snippets or analyzing and optimizing existing code.
These hurdles are especially high for inexperienced developers. They take a long time to become productive and can't fully tap their potential. This leaves efficiency reserves in teams unused. The industry is evolving, though. Competitors like Siemens already use smart AI assistants to support their developers. Modern solutions like GitHub Copilot and Amazon CodeWhisperer show how AI models can help programmers not just with knowledge transfer, but directly in coding.
Target Vision
Together with WAGO, we developed a target vision for an AI-powered assistant that optimally supports PLC programmers in their daily work. The assistant should serve as a central tool that provides not only WAGO product-specific knowledge and a structured knowledge base, but also precise, context-related answers. The goal is for the assistant to make code suggestions on its own, analyze existing code, and proactively help solve complex problems.
A key feature of the PLC Agent is its ability to generate IEC code specifically for connecting WAGO hardware. If a developer wants to address a specific WAGO terminal via IEC code, for example, the agent creates the matching source code, precisely tailored to the specifics of the hardware being integrated. This takes the PLC Agent far beyond a classic code generator, as it consistently aligns with the concrete requirements and properties of the respective WAGO component. Operation happens via a web interface or a custom CODESYS add-on to ensure the greatest possible integration into the development workflow.
The PLC Agent will be integrated into WAGO’s IT landscape to make efficient use of existing resources and continuously access up-to-date knowledge. Through this connection, the PLC Agent will be continually enriched with current information, ensuring developers always have access to the latest knowledge.
Our Services
The WAGO PLC Agent was developed in three service areas:
- Consult: Conceptualization of a future-proof AI system aimed at sustainable integration into the existing IT landscape.
- Develop: Build of the AI assistant with specialized WAGO knowledge, introduction of a QA system for continuous monitoring of user satisfaction and response quality, CODESYS integration.
- Services: Integration with internal company systems for automatic updates and ongoing optimization based on feedback and new requirements.
Customer Benefits
The WAGO PLC Agent brings immediate advantages to daily developer work. Programmers get instant answers to their questions without having to search through documentation for long. This saves time and significantly speeds up development projects. It's especially worth noting that the PLC Agent generates IEC code directly for the specific WAGO hardware and automatically reads the project-specific context. This way, the PLC Agent adapts the code and suggestions to the respective project and development context. The PLC Agent's capabilities thus go beyond those of a classic code generator and enable integrated, efficient, and hardware-oriented development.
The entry barrier for new developers drops noticeably. Programmers with less experience can take on more complex tasks because the company's specialized knowledge is always available. This means faster productivity and more flexible staffing. For WAGO as a company, it opens a new business area. A scalable service offering positions WAGO as an innovative technology provider. Customers benefit from a continuously improving service.
Early demonstrations already showed concrete successes. The assistant successfully generated code for practical tasks, for example programming specialized measurement terminals, and explained error codes.
Our Expertise
The system draws on several specialized areas:
- Artificial Intelligence and Language Models: Use of modern AI technology to enable natural communication between human and machine.
- Systematic Knowledge Integration: Connection to source systems and creation of a central data platform including data pre-processing and vectorization.
- Cloud Infrastructure: Secure provision and scaling in the WAGO cloud for reliable operation and data protection.
- Integration into Development Environments: Incorporation of the assistant into developers' existing workflows.
- Domain Knowledge: Deep familiarity with IEC programming, WAGO products, and engineering practices.
- Quality Assurance: Continuous improvement of response quality, ensuring reliability and ethics standards.