Smart marble cutting through a vision system

Abstract The paper describes a vision system equipping a CNC machine tool for cutting marble and similar stones, which allows a semi-automatic detection of surface defects and then computes an optimal distribution of the workpieces to cut by avoiding the defective parts of the slab. In order to address the workpiece allocation, a nesting algorithm is exploted which solves the so-called irregular packing problem with a limited computational time by largely reducing the time required to prepare the cutting program. The system has been fully implemented and tested on a real CNC machine tool of a marble manufacturer: the experimental setup and some exemplar results of the workpiece allocation procedure are described and discussed.