Bubbling is commonly used in the chemical and mineral industries to enhance heat and mass transfer in reaction vessels. However, it is difficult to quantify the bubbling effect directly and reliably, thus its control is dominated by manual procedures. In the case of bottom bubbling, the bubbling often creates a disturbance at the top of the vessel that can be observed. For example, in ladle metallurgy, the bubbling in the vessel creates a spout eye at the surface that is directly related to the rate of bubbling, the physical properties of fluids and geometry of the system. This phenomenon is accompanied by sound and vibration signals that are closely related to each other. In practice, operators make judgements about the process through visual observation, listening and feeling vibrations. Therefore, these signals provide information about the bubbling phenomena, which could be analysed mathematically to provide a quantitative signal suitable for controlling the process. In this study, we have gathered images, sound and vibration data from bubbling experiments, established techniques for analysing the combined data matrix using PCA (Principal Component Analysis) method to identify latent variables that maybe suitable for online control of the process. Our results show that one or two latent variables formed from the combined signals do capture the great majority of the variation in the system and that simultaneous online analysis of bubbling phenomena is feasible