Our dry, zero-pollution laser-based metal coloring process addresses the limitations of conventional laser coloring methods in color gamut and precision. It integrates sub-pixel color mixing to expand achievable colors, and employs a physics-driven deep learning model (PD-TNN) with thermal simulation to map process parameters to color outcomes, enabling accurate prediction and efficient process design.

Our technology addresses challenges in metal laser coloring—narrow color gamut, trial-and-error parameter design, and poor inverse prediction—through multiple innovations. These include a PD-TNN model for accurate inverse mapping, sub-pixel mixing for enhanced color expression, and thermal simulation for establishing a parameter–temperature–color database. By integrating AI, process design, and simulation, this approach enables a practical and scalable coloring method.

Our technology is a high-efficiency, low-carbon laser metal coloring process that replaces wet methods like electroplating, painting, and PVD. Using room-temperature dry processing aligned with ESG trends, it integrates AI for color prediction and parameter control. The system supports intelligent, customizable production and has applications in both industry and design, targeting commercialization in Taiwan’s surface treatment market, which exceeds NT$100 billion in scale.