Progress measured,
not estimated.

AI construction progress tracking means comparing site photos, 360 captures or laser scans with the BIM model and the programme, so the percentage complete for each area and activity is measured from evidence rather than estimated, and slippage on the critical path shows within the week.

Reality-capture platforms already turn a walk with a 360 camera into a navigable record, and some compare it with the model. The harder part is the link to the programme. Activities in Primavera P6 or MS Project are rarely broken down the same way as the model, so a wall recognised as built does not automatically move any activity forward. Without that mapping, progress tracking produces good pictures and the same Friday estimate.

So most of the build is the mapping: model elements and zones to programme activities, agreed with your planner, and then a weekly comparison that proposes progress per activity with the images that support it.

We run it alongside your current reporting for several weeks on one site. The proposed progress is compared with what the site team confirms, activity by activity, and the differences are reviewed with them. That tells you which trades the image recognition handles well, usually structure and envelope, and which, such as concealed services, still need a person to report.

Site imagery shows people, so it is personal data under the GDPR. Capture routines are planned to avoid faces where possible, images are stored inside your environment, retention is set and documented, and workers are told how imagery is used. Drone capture in the EU falls under Regulation 2019/947 and its national rules, which the operator must follow. Those decisions and duties are yours; we build the controls that support them.