In short
- Good candidates are the documentation tasks: reports and expert opinions from plans, photos and measurement data, taking quantities off drawings, writing site reports, maintaining technical documentation and data sheets, compiling project status from several systems.
- Your programs stay. The AI reads plans as PDF, IFC or DWG and writes through the interfaces of your project and construction software, a new system is not needed.
- Professional responsibility stays with the person who signs the report. The AI delivers the draft, and every figure names its source in the plan.
- Where images, plans or measurements are unclear, the AI flags it and does not guess.
AI in project management: seven tasks at a glance
These are typical workflows, not client projects. In each one the AI prepares, and a professional checks and decides. An example of the first workflow is shown on the page on AI automation, and documents are covered in more depth on the page on document processing.
1. Reports and expert opinions from plans, photos and measurement data
A technical report summarises what is in plans, on photos and in test records, and takes days to write.
- Input: plans and revision states, photos of the site visit, measurement data and your report template.
- AI prepares: reads the documents, assigns statements to the chapters of the template and writes a draft. Every statement names plan, photo or measured value as its source. Unclear passages, such as a blurred photo, are flagged.
- Professional checks: the draft sentence by sentence and adds what can only be seen on site.
- Output: a report draft in your template, with references to sources.
2. Quantities and costs from plans
Quantities from drawings are often counted and measured by hand, and the effort grows with every change to the plans.
- Input: plans as PDF, IFC or DWG, bill of quantities and price list.
- AI prepares: reads components, areas, lengths and counts from the plans, suggests items and calculates the totals with code. Prices come from your price list.
- Professional checks: the quantities against the plan and decides what goes into the quote or the final account.
- Output: a quantity schedule with the source in the plan, a cost draft with prices from your list.
3. Site reports and daily logs
Daily reports are written in the evening, from memory and from notes.
- Input: voice note, keywords, photos and the day's time records.
- AI prepares: writes the daily report from them with weather, staff, work done, delays and special points and assigns photos.
- Site manager checks: the report briefly and corrects.
- Output: a site report in the agreed format, filed in the project file.
4. Check tenders and contracts
Before submission or signature, requirements, deadlines and special points have to be found.
- Input: tender documents, contracts and technical requirements.
- AI prepares: summarises requirements, dates, contractual penalties and warranty and compares them with your standards. Differences are shown with the passage.
- Professional checks: the summary and reads the critical passages personally. Legal assessment is for lawyers.
- Output: a short overview with passages and a list of differences.
5. Technical documentation and data sheets
Documentation tends to be written at the end, when everyone is already on the next project.
- Input: design data, test records, manufacturer information and earlier documentation.
- AI prepares: assembles the documentation from the documents, spots missing details and matches data sheets with the project state.
- Professional checks: content and completeness. Safety-relevant details the professional checks personally at the source.
- Output: documentation in the format the client or authority requires, with open points.
6. Project status from several systems
The state of a project is spread across plans, time records, procurement and emails.
- Input: data from project software, time tracking, ERP and mailbox.
- AI prepares: code calculates progress, hours and costs against plan. The AI writes the status report from them and names the figures it relies on.
- Project manager checks: the report and adds what the figures do not show.
- Output: a status report with differences and open points.
7. Timesheets
Hours are recorded late and imprecisely, and assigning them to projects takes time.
- Input: voice notes, calendar, daily reports and time tracking.
- AI prepares: suggests the assignment of hours to projects and activities and reports gaps.
- Employee confirms: their own hours. Nobody is rated or compared by the AI.
- Output: a confirmed timesheet for time tracking and invoicing.
Programs and file formats: connected to what you already use
In engineering and projects the data sits in plans, spreadsheets and specialist programs. The table lists the usual cases. Which route your system offers, we clarify before the quote.
| Program | What is exchanged | Route |
|---|---|---|
| Project management (for example Jira, Asana, Microsoft Project) | Tasks, dates, status | Interface (API) of the program or export |
| Tendering and construction software | Bills of quantities, quantities, invoicing | Import and export formats of the program |
| CAD and BIM | Plans, components, quantities | Read the files: PDF, IFC and DWG as input |
| Document management | Plans, reports, contracts | Filing with a reference in the project file |
| Time tracking | Hours per project and activity | Interface or export |
| Excel calculation | Prices, tiers, calculations | Read the files, source for every cell |
The names are for orientation and do not mean a partnership. Where a program offers no interface, we work with export and import, a watched folder or email.
Connect, do not replace
You do not need new project software or new construction software, and we do not sell any. The AI works next to your programs: it reads plans and documents and writes drafts.
- Your programs stay in charge. Schedules, calculations and invoices live there. The AI writes only into fields you release.
- Your report formats stay. Reports look as before and are ready sooner.
- One workflow at a time. We start with one step, for example the daily reports, and run it alongside the existing work first.
- The way back stays open. All documents live in your systems. Without the AI your people carry on as before.
Approvals and responsibility
In engineering, whoever signs is liable. So before the start we agree what the AI may do.
- Professional responsibility: It stays with the engineer, expert or project manager. The AI delivers a draft, not an assessment.
- A source for every figure: Every quantity, dimension and statement names plan, page or measured value. What cannot be supported is not in the draft.
- Review note: Before a report goes out, the reviewing person notes their approval.
- Unclear passages: Blurred photos, contradictory plans or missing measurements are flagged and not filled in.
- Safety-relevant points: Details on structure, fire protection, electrics and hazardous substances are checked by the professional personally at the source.
- Log: Every draft is recorded with sources, time and reviewer.
Who is liable when an AI draft contains an error cannot be answered across the board. As a rule the person who approves the report remains responsible, as with a spreadsheet or a drawing prepared by an assistant. How your professional indemnity insurance and your professional body view the use of AI drafts, please clarify before use. Where data of employees is touched, for example on timesheets, technical systems need the involvement of the works council in Germany under section 87 BetrVG and in Austria under section 96a ArbVG. This overview is general information, not legal advice.
Data: plans, photos and measurement data
The quality of the drafts depends on the documents the AI works with.
- Plan states and revisions: The AI has to know which state applies. We read plan index and revision details and report when states contradict each other.
- Photo documentation: Photos need date, place and a link to the component. We assign them from metadata and notes and flag what we cannot assign.
- Measurement data: Measured values come from records and files of the measuring devices. Code calculates with them, the AI describes.
- Project file and access: Where the documents are and who may access them is for you to decide. The AI sees only what it needs for the step.
- Data protection: Photos of building sites often show people, and reports name names. We process everything on servers in the EU and agree retention and deletion with you. More under AI and data protection.
Your own apps for office and site
Some things are missing not in the project software but on the smartphone on site. We build them as separate programs when standard products do not fit:
- Measurement app: record dimensions and photos, see quantities at once, hand the result to the calculation.
- Site diary app: voice note and photos of the day, from which the daily report is made.
- Defects app: record defects with photo, place and deadline and track them until they are fixed.
- Digital timesheet: confirm hours on the phone, assigned to project and activity.
These programs complement your project software, they do not replace it. More under app development and software development.
Where to start
These three questions help with the assessment:
- Which report costs your engineers the most time, and what is it made from (plans, photos, measurement data)?
- Which project software and which formats do you use, and where do plans and project files sit?
- How do daily reports and timesheets come about today, and when?
If you send us the answers, we tell you which workflow is worth doing first and what has to stay with your professionals. We begin with a trial run alongside the existing work.
Sources
- EU: Regulation (EU) 2016/679 (GDPR), art. 28: EUR-Lex; Regulation (EU) 2024/1689 (AI Act), Annex III: EUR-Lex
- Germany: section 87 BetrVG
- Austria: section 96a ArbVG