Digitalisation
AI in construction
Where artificial intelligence helps in construction today, what it needs and how to set up a sensible first pilot.
Where AI helps today
In construction AI is strongest where many similar documents, images or data arise and people have so far sorted, searched or checked by hand. An overview by project phase:
| Phase | Use case | What it needs |
|---|---|---|
| Design | Check design documents against regulations (automated compliance checking) | Machine-readable rules and structured models |
| Tendering | Classify and compare bill-of-quantities items | Consistently structured bills of quantities |
| Construction | Capture progress from photos, pre-sort defects | Photos with place and time, a reference model |
| Control | Spot schedule risks early from progress data | Maintained schedules across several projects |
| Office | Open up knowledge and documents, pre-answer requests | Filed, findable documents |
We do research on automated compliance checking1 ourselves, see our publications.
Placing use cases
Not every use case pays off first. A simple grid of benefit and effort helps to start with a quick win and to prepare the big projects well.
In our own office
Two digital colleagues work with us: Rheo answers in the team chat and takes on small tasks, Onda takes calls when nobody is free and passes on the request with a transcript. We introduce both on the team page.
Before the first pilot
- Pick a concrete problem that costs measurable time today.
- Check whether the data needed exists, is accessible and may be used.
- Clarify who checks and answers for the result.
- Define success up front, such as hours saved per month.
- Start small, evaluate, then scale up.