Estimating attached mortar paste on the surface of recycled aggregates based on deep learning and mineralogical models

•Image analysis of recycled aggregates as a manual procedure to estimate the surface of attached mortar content.•Artificial Intelligence applied to the automatic classification of recycled aggregates.•Bulk volume reconstruction of attached mortar based on X-ray Powder Diffraction mineralogical model.•A direct method based on a mineralogical model for determining the extent to which the attachedmortar is carbonated.•Tailored final applications based on the attached mortar content and combined carbonation degree.