The MAPDRONE project, funded by Multiprojectus and developed in the field of Civil Engineering, has as its main objective the creation of an intelligent methodology for the detection and mapping of anomalies in the external envelope of industrial buildings, integrating 3D photogrammetric models with artificial intelligence.

By combining advanced 3D reconstruction techniques with computer vision algorithms (such as Mask R-CNN), MAPDRONE aims to provide an automated, scalable, and accurate approach to remote technical inspections, reducing costs and increasing reliability in the assessment of industrial façade conditions.

The main objectives of MAPDRONE are:

• Systematization of procedures and best practices for the development of three-dimensional photogrammetric models through the fusion of digital technologies.
• Development of an Artificial Intelligence algorithm (Mask R-CNN) for anomaly segmentation (corrosion, mechanical impact, and water accumulation) in sandwich panels.
• Application of the anomaly mapping methodology, based on 3D photogrammetric models, to the case study of an industrial building.