Predictive Quality Control / solino Application note J^)tO100-SO-10-MAN-001 Watch Movement Anomaly Opto solino RTI sensorDetecting micro defects using BRDF computational imaging Customer Task: Precision watch movements contain dozens of miniature components made from highly polished or decorated metal surfaces. They must meet extremely high-quality standards because even the smallest surface defect can affect functionality, reliability, and the perceived value of the final watch. • jewel bearings • decorative finishes The goal is to detect surface defects early in the production process and ensure consistent quality across all assembled movements. Inspection Challenges: Inspecting inner watch mechanics is particularly challenging for traditional machine vision systems. Typical challenges include: • Highly reflective metal surfaces • Complex geometries and small components • Decorative finishes that change appearance with lighting • Micron-scale scratches or dents • Particles or contamination inside the movement Standard camera systems rely on fixed lighting conditions and often fail to reveal subtle surface irregularities. Small defects may only become visible when the surface is tilted under specific lighting angles. As a result, many inspections still rely on manual visual inspection performed by skilled operators. Fig. 2: 100-SO-10-MAN-001 solino 10x10 - Viewer - Stand man Opto GmbH • Floriansbogen 2/4 • 82061 Neuried • Germany Tel: +49 89 8980 55 0 • [email protected]

Technology Principle The solino RTI sensor uses BRDF-based computational imaging to reveal surface structures. Within milliseconds the system captures a series of images under different illumination angles and reconstructs the surface reflectance information computationally. solino 10x10 is used to inspect assembled watch movements. Typical inspection targets include: • gear wheels, pinions or bridge surfaces • screws and screw heads • decorative surface finishes • contact surfaces between moving parts • photometric stereo image capture • reflectance-based surface reconstruction • computational...