Computer vision
for construction

On a large construction site, deviations from the design are found by technical-supervision engineers: it is expensive, slow and subjective. Checking against the BIM model took 2-3 days and held up the acceptance of stages, while the bank tranches depended on an objective readiness figure.

• shorten the cycle of control and BIM comparison
• give the bank an objective readiness figure
• stop arguing with contractors by eye
• cover safety control
• not stop the work on site during surveys

Comprehensive automation of quality control:

• AI + computer vision + IoT
• Equipment: 360° cameras, lidars, thermal imagers, humidity sensors
• High-speed shooting: 1 minute per room
• AI analysis and automatic defect detection: in 1 day
• Comparison of actual conditions with BIM models
• Web panel and mobile applications for operators
• Integration with 1C and customer systems