Appendix A — Tools
Google Colab in five minutes, and AI assistants
After this page
- Notebook open in Colab from the week page
- Own copy on Google Drive
- Back on track after Colab forgets everything
What Colab is
Google Colab is a notebook in your browser: text cells and code cells, and every code cell runs on a Google computer, not on yours. Think of a lab bench Google lends you for the afternoon: the tools are laid out, and when you leave, the bench is cleared. Your notebook stays on your Google Drive; the computer’s memory does not. A free Google account is enough; nothing to install.
Four steps, once per week
- Open. Click Open … in Colab on the week page. (In class the notebook is also on ILIAS: download it, then File → Upload notebook.)
- Save a copy. File → Save a copy in Drive. Work in your copy; the original stays clean.
- Run the first code cell. It loads the tools. Warning “not written by Google”: click Run anyway.
- Top to bottom. Every step ends with a check, ✅ or ❌. A ❌ comes with a hint: fix one line, run again.
When something goes wrong
| What you see | What it means | What to do |
|---|---|---|
NameError: name '…' is not defined |
A cell above was not run since Colab woke up | Runtime → Run before, then the cell again |
| “Session crashed” / everything grey | Colab’s computer was recycled | Runtime → Restart session, run from the top |
| A check shows ❌ | Your number differs from the expected one | Read the hint, compare with your hand calculation |
| Your copy is gone | It was not saved to Drive | Open from the week page again, save a copy first |
AI assistants
Allowed in every task. Two habits: predict the answer before you ask, and check the assistant’s number against one you computed by hand.
- Academic Cloud — the service portal of German universities, run by GWDG (Göttingen). Sign in with your TH Köln account and use several language models in Chat AI, free of charge.
- Prompts for this module — quiz-me prompt for revision with the week’s key content.
- How to work with an assistant — AI Fluency: the four Ds, which model for which task, two applied examples.
Principles for GenAI in this programme
- Support, not a substitute for thinking. Using GenAI to support your own work is welcome and recommended. It must not replace your own academic work.
- Be able to explain it yourself. At any time, you must be able to explain your own work orally, on your own and in answer to follow-up questions.
- Avoid your own work, and you fail. If central parts were not produced by you, the result is a fail (example: whoever cannot coherently explain a source or method more than twice has, as a rule, failed the examination).
Translated from the German original: Bartnik, R., & Haag, C. (2025). Rahmensetzung zur Nutzung generativer KI im Bachelor- und Master-Studiengang Wirtschaftsingenieurwesen (V02). TH Köln. PDF