In the search for smarter and more efficient welding methods, the measurement of the quality of parts is fundamental but limited since, traditionally, it is necessary to destroy samples of a prodution batch to estimate this quality. However, for an online estimation and possible corrections in production, this dataset brings the data result of an attempt to estimate the width and reinforcement of the weld bead through acquisition techniques with thermographic sensors and profilometers on a motorized table and controlled in a GMAW process.