The next leap in industrial-grade AI

24 November 2025  

The Industrial Foundation Model understands text, 3D models and other complex structures. It could boost productivity and efficiency in automation and engineering.

Gen AI can read text and images, but struggles with complex industrial data. What if AI could be trained to understand the language of engineering and manufacturing? Koushik Kasturi, a product marketing coordinator in Siemens Digital Industries Software, explains how the company is developing an Industrial Foundation Model that brings true industrial intelligence to manufacturing.

In 2022, OpenAI shook the world by introducing ChatGPT. Since then, generative AI (GenAI) has begun supporting tasks in engineering and manufacturing such as generating technical documentation, boilerplate programming, summarising data, and answering common technical questions.

But here’s the big question: If you work in engineering or manufacturing, is today’s GenAI enough for you?

The short answer is: No.

The industrial world is far more complex and runs on 3D CAD data, bills of materials, process simulations, toolpaths for CNC programming, tolerances, and quality inspection datasets. If AI can’t interpret these, it can’t truly collaborate with engineers and manufacturers.

That’s why we’re building something different – an AI that understands your domain from the ground up.

We are entering a new era where AI is not just a generalist. It becomes a specialist that understands the precision of engineering data, the complexity of manufacturing workflows, and the standards that keep operations running smoothly.

To make AI truly industrial grade, it must be trained to speak the language of engineering fluently.

The Industrial Foundation Model is a large-scale generative AI fine-tuned for the industrial world. This means it truly understands the language of engineering and can work with the complex types of data that engineers handle every day.

Trained on vast amounts of industrial data, it can deliver precise and relevant answers tailored to daily engineering and manufacturing needs. Built to meet industry standards, this model offers the trustworthiness, accuracy, reliability, and security that are essential for real-world industrial applications.

AI is transforming how products are designed, planned, and produced. By combining data, automation, and intelligent insights, AI helps manufacturers to work faster, make better decisions, and deliver higher-quality results

AI-assisted CAM programming

Ask any NC programmer and they will tell you that too much time is lost in repetitive, manual toolpath creation. CAM programmers spend too much time generating and adjusting the toolpaths needed to manufacture components. Gaining the necessary expertise takes years, making it costly and reliant on a limited skilled workforce. The Industrial Foundation Model can streamline this process by understanding engineering data, industry best practices, and knowledge from past projects.

The AI-powered model can suggest cutting strategies, tools and toolpaths, while enabling direct interaction via an intuitive user interface. It can identify machining features, recommend strategies, and justify decisions, automating key aspects of CAM programming to save time and reduce skill dependencies.

In Siemens’ cloud-based NX X Manufacturing CAD/CAM tool, an AI co-pilot instantly reads the geometry, identifies features, and delivers the right cutting strategy, tooling and operations. No library searches or repeated data entry. Just fast, intelligent, and context-aware programming.

AI-powered manufacturing planning

Planning engineers spend a lot of time turning design data into detailed manufacturing plans. Tasks such as converting an EBOM into an MBOM, or a bill of process, can take hours or even days. The Industrial Foundation Model is can change that by connecting directly to robotic simulators and factory layout tools.

Using this AI technology, planning steps that previously took forever – such as generating production line layouts that follow company-specific rules or running virtual process validations – happen quickly and automatically. This means less time is spent on repetitive work and more time for engineers to focus on solving problems, improving workflows, and driving innovation.

Simulation and validation processes have always required a careful balance between accuracy, time, and resources. Engineers often spend hours creating datasets, preparing separate models, and running lengthy validations before a single part is produced. The Industrial Foundation Model is set to change this.

By working directly with connected process planning datasets, AI can validate assembly processes in 3D on the fly without needing to duplicate data. It can check feasibility, highlight potential issues, and suggest process improvements well before production begins.

Factory layout planning also benefits because AI can automatically factor in all company-specific regulations and requirements, ensuring compliance without manual effort.

Generative AI brings many benefits, but through conversations with customers, three key advantages of the Industrial Foundation Model stand out:

• Bridging the skills gap  By gathering and organising company and industry knowledge in one place, the Industrial Foundation Model makes it easier for new engineers to learn. New recruits can quickly understand processes, standards, and best practices, without spending months shadowing or searching for information. This helps close the skills gap and eases pressure caused by the shortage of experts.
• Enforcing standards  Trained on company-specific data and industry best practices, the model helps engineers to follow the right standards and avoid costly mistakes. This ensures work stays consistent with company guidelines, reducing errors and rework.
• Boosting productivity  By taking on repetitive, manual, and non-innovative tasks, the Industrial Foundation Model frees experienced engineers to focus on what really matters – solving complex problems, exploring new ideas, and applying their creativity. This means they spend less time on routine work and more time driving meaningful improvements, allowing skills to be used where they have the greatest impact.

The road ahead

The Industrial Foundation Model is designed not only to understand engineering data but also to think through engineering challenges. It learns the data, rules, and precision that drive industrial processes. The model will transform how companies design, plan, manufacture, and operate across the entire value chain.

Human–machine collaboration in industry is no longer just a concept — it is happening right now.