1：Experimental Design and Method Selection:

The system consists of an industrial robot, a scanning head with optical triangulation feedback, and a fiber laser. The methodology involves detecting the keyhole and its shape and size on the acquired image, extrapolating the detected seam curve to the nearest point relative to the keyhole, and calculating the new laser position and power.

2：Sample Selection and Data Sources:

Commercially available 304 stainless steel plates with dimensions 300 mm x 30 mm were used for validation purposes.

3：List of Experimental Equipment and Materials:

Industrial robot (Yaskawa MC2000), scanning head (HighYag, RLSK) with a console attached camera (PointGrey, Flea3, model FL3-U3-13Y3M-C), fiber laser (IPG, YRL-400-AC, 400 W), illumination laser (Fotona XD-2, 810 nm ± 10 nm), bandpass filter (Thorlabs, FBH810-10).

4：0). Experimental Procedures and Operational Workflow:

4. Experimental Procedures and Operational Workflow: The system was tested on a curved flat edge seam and a 2.0 mm thick lap joint. The welding velocity was set to 100 cm/min for seam tracking validation and 60 cm/min for power control validation. Camera frame rate was set to 80 frames per second.

5：0 mm thick lap joint. The welding velocity was set to 100 cm/min for seam tracking validation and 60 cm/min for power control validation. Camera frame rate was set to 80 frames per second. Data Analysis Methods:

5. Data Analysis Methods: The keyhole location and shape were determined using image processing techniques, including Gaussian filtering and morphological operations. A PI controller was used for laser power control.