Automation and artificial intelligence in structural engineering
Discuss engineering automation, Python, BIM workflows, parametric modelling and the verification of AI-assisted work.
Start with a defined engineering task
Automation is easier to evaluate when the task is specific. Describe the input, expected output, software environment and review responsibility. Examples include checking a model export, preparing repetitive drawings, comparing analysis results or connecting a calculation workflow to BIM data.
Make the workflow reproducible
For Python scripts, spreadsheets, APIs and parametric models, include a small reproducible example with units and expected behaviour. Record software and library versions when they matter. Explain which parts are deterministic and where a person must interpret or approve the result.
Discuss AI with evidence
For an AI-assisted workflow, distinguish generated suggestions from verified engineering information. Useful conversations describe how references were checked, how unsupported output was detected and how the workflow handles confidential information. Share a review method or a bounded experiment rather than an unverified claim of accuracy.
Questions to bring to the community
- Which repetitive engineering task would you automate first?
- How do you test a script against a known calculation?
- What evidence do you require before accepting AI-generated output?