Course Catalogue.

After participating in this course, you will be able to:

• Understand what AI foundation models and LLMs are, as well as their application in engineering
• Effectively use AI within engineering workflows (design, documentation, programming, simulation support)
• Recognize the limitations, risks, and best practices related to the use of LLMs
• Experiment with real-life cases and practical tools

Description
AI foundation models such as LLMs are experiencing rapid advancement and are now being integrated into numerous fields, including engineering. In the face of this constant evolution, it is essential to stay informed about the latest developments, understand how these tools can be utilized, and recognize their limitations. This way, engineering professionals will be better equipped to use these technologies wisely and effectively.

This course emphasizes the practical application of LLMs within engineering workflows. Participants will learn to leverage these AI models to support design, generate technical documentation, automate programming tasks, facilitate simulation, and solve complex problems. Through hands-on cases and interactive demonstrations, participants will discover how these tools can optimize processes, increase productivity, and drive innovation within engineering teams. Exploring both the strengths and limitations of AI for engineering will also enable the integration of best practices and help manage the risks associated with its everyday use.

Who Should Attend
Engineers, engineering managers, technical consultants, technicians, technologists and researchers with little or moderate exposure to artificial intelligence (AI)/LLMs within the engineering field.

Course Syllabus

Module 1: Introduction to LLMs and Foundation Models (1h)

• What is a foundation model? What is an LLM?
• Evolution: machine learning → deep learning → transformers → LLM.
• Key concepts: tokens, embeddings, context windows.
• Overview of available models: GPT, Claude, Gemini, open-source models such as Llama, Mistral.
• Applications in engineering: documentation, design support, simulation, troubleshooting, project planning.
• Interactive demonstration: Ask ChatGPT/Claude to explain an engineering concept at different levels (e.g., fluid mechanics for a 10-year-old versus for a PhD student).

Module 2: LLMs in Engineering Workflows (1h30)

Use Cases:

• Writing technical reports and standards.
• Programming assistance (MATLAB, Python, C++ for simulations).
• Knowledge management (summarizing scientific articles, extracting design data).
• Troubleshooting and Q&A for engineering systems.
• Generating training materials and safety manuals.
• Basics of Prompt Engineering: role, few-shot prompting, structuring thoughts.

Exercise: Participants create prompts to:

• Generate a troubleshooting checklist for a hydraulic pump.
• Summarize a scientific abstract in 3 points.
• Translate a safety procedure into plain language.

Module 3: Practical Workshops (2 hours)

Group work with LLM tools.

Workshop 1 – Documentation & Automation:

• Use an LLM to generate a project report template or a design review checklist.

Workshop 2 – Simulation & Programming Assistance:

• Use an LLM to:
• Write a Python function to calculate stress/strain.
• Debug a MATLAB script for a control system.

Workshop 3 – Knowledge Extraction:

• Load a scientific article or manual → extract the main design parameters with an LLM.

Module 4: Limitations, Risks, and Reliability (1 hour)

• Limitations: Hallucinations, outdated knowledge, lack of domain expertise.
• Bias & Ethics: Ethical considerations and bias in engineering decision-making.
• Validation and Verification: When not to trust the model.
• Data Confidentiality and Intellectual Property: Protecting sensitive information and IP rights.

Exercise: Present an incorrect calculation generated by an LLM. Participants must identify the errors and discuss methods to validate results.

Module 5: The Future of LLMs in Engineering (45 min)

• Integration with engineering software (CAD, simulation tools).
• Intelligent assistants for design, maintenance, and operations.
• Multimodal models (text + images + diagrams).
• Autonomous agents in engineering workflows.

Discussion: “Which part of your engineering work could truly be improved by Co-Pilot AI?”

Adjournment & Summary (30 min)