Another sector for which AI services have been developed is infrastructure construction. These AI services are mainly used to recognise damage in the infrastructure in images. Segmentation can be used to classify and precisely localise damage such as spalling, spalling with exposed corroded reinforcement, gravel pockets, gravel pockets with exposed corroded reinforcement, cracks and corrosion. This service enables the first steps towards the automation of building inspections. On the one hand, the documentation of damage is automated. Secondly, the damage in the structure is localised precisely, making it easier to find damage in future inspections and to record the development of damage.

3D modeling in AI services for infrastructure construction

Infrastructure also includes the reconstruction of bridges. Construction plans serve as the basis for this. An AI service extracts the bridge gradients from the longitudinal view. This combines deep learning-based detection with optical text recognition to obtain information about the height and inclination of certain points. Together with other AI services, the 3D position of the structure’s axis can thus be described. This enables the correct 3D modeling of bridges.