A calibration error in a diamond mine caused high-value gems to be discarded as waste. The technical expert report revealed that the conveyor belt speed generated a motion blur effect in the X-ray data, distorting the volumetric reconstruction. This case demonstrates how process simulation allows identifying invisible faults in real operation.
Volumetric reconstruction and detection of kinetic artifacts 💎
The sorting system uses Industrial CT Software to reconstruct the internal shape of each mineral fragment from X-ray projections. When the belt moves at a constant speed, the object shifts during acquisition, generating a blurring artifact in the tomographic slices. MATLAB processes the temporal signals and applies deconvolution filters to quantify the actual displacement. Unity 3D recreates the material flow in a virtual simulation, allowing parameters such as belt speed and detector exposure time to be adjusted without stopping production.
Validation of corrections before real implementation ⚙️
The simulation in Unity 3D allows testing corrections such as synchronizing the X-ray beam trigger with the exact position of the object. By modifying the belt speed or acquisition frequency in the virtual environment, it is observed how the motion blur disappears and the reconstruction recovers the actual geometry of the gems. This methodology reduces costs and avoids unnecessary stoppages in the production line, demonstrating that a calibration fault can be resolved by analyzing the entire process from the simulation software.
How were the discrepancies in the sorting sensor calibration detected during the 3D simulation of the mineral classification process?
(PS: Simulating industrial processes is like watching an ant in a maze, but more expensive.)