AI release promises construction quality control leap

Dashboard image of new cloud-based 3D AI Construction Inspection system

German engineering services company TÜV SÜD has partnered with UK construction analytics firm Contilio to develop a quality control system based on artificial intelligence.

The system, 3D AI Construction Inspection, provides automated defect analysis and quality verification for building and infrastructure projects.

TÜV SÜD says that, while technologies such as BIM (building information modelling) are making huge strides in improving productivity and sustainability in the construction process, their impact is often negated when advanced technology is not used to keep them up-to-date and compliant throughout the project’s cycle.

The new system will support risk-management and other project decisions and give all stakeholders access to the same information from a cloud-based environment.


Intelligence
Zara Riahi, CEO of Contilio, said, “Our solution combines advanced 3D AI technology with 150 years of global inspection and testing expertise to automatically provide actionable intelligence from 3D construction site data (Lidar/Laser scans) in real-time. This is done by intelligently identifying all structural, architectural, MEP and fitout elements in 3D site data, and then automatically comparing them with the BIM and the schedule.”

Riahi added, “Our AI software works seamlessly with commercially available Lidar scanners and drones and helps to automatically track construction progress and installation quality for both building and infrastructure projects. In addition, this automated comparison helps keep digital twins up-to-date during the construction and provides the opportunity to rapidly detect any deviations from plans or installation errors and eliminate them with agility.”

Joachim Birnthaler, CEO of TÜV SÜD’s Real Estate & Infrastructure Division, said, “Shockingly, 60% of all construction projects experience significant delays, and up to 30% of construction costs are swallowed up by the need for rework and installation error rectification.”