The laser beam welding technology provides high exibility and adaptability to the welding process, accommodating complex joint geometries efectively. Furthermore, using laser sources mitigates microstructural alterations in materials attributed to the technology's low heat input and minimal thermal distortion. Consequently, the adoption of laser beam welding technology is prevalent in sophisticated markets, notably electric vehicle battery manufacturing. Nevertheless, these emerging application domains pose challenges necessitating resolution, such as joining dissimilar materials characterized by varying melting points and high reectivity values at the wavelengths of the most common high-power industrial lasers (e.g., aluminum and copper).

This Ph.D. thesis presents an adaptive control method for joining high reectivity dissimilar materials using spatial and temporal modulation of the laser power intensity for a dual laser beam setup to overcome the challenges of welding this kind of material combinations. The application of deep learning models for classifying the resulting welding strategies is also explored for automated process monitoring based on high-speed infrared imaging. As a result, a comparative study of diferent combinations of pulsed and continuous wave laser arrangements through static and dynamic optics and the obtained joining qualities is presented.