Noise control is an integral part of acoustics engineering and a concern for modern society. For example, the EU Action Plan: 'Towards Zero Pollution for Air, Water, and Soil' aims to reduce the share of people chronically disturbed by transport noise by 30%. However, wave control is normally constrained by the physical and geometrical properties of the materials. The recent concept of active metasurfaces made from acoustic metamaterials is opening new horizons for sound control applications, as the properties of these metasurfaces can be modified to achieve the desired acoustic behaviour and can be tailored to specific applications. A metasurface that could adapt itself both in space and time, i.e., spatiotemporal modulation, based on the wavefield could open new possibilities for controlling sound in real time. However, implementing this type of metasurface still presents significant challenges due to experimental difficulties and control algorithm limitations. This project addresses these challenges by proposing an acoustic metasurface composed of individual electroacoustic absorbers that can be controlled with a pressure-current-based control system. Therefore, INTSURFACE research aims to develop a real-time adaptive metasurface with spatiotemporal modulation capabilities. During this project, the theoretical and numerical frameworks to evaluate metasurfaces will be expanded to include space-time dependent properties and allow for the prediction of their acoustic behaviour in practical scenarios. Then, to control the metasurface, the research will develop a machine learning control framework that can adapt the individual elements of the metasurface to implement a desired spatiotemporal configuration. Finally, a proof-of-concept acoustic metasurface will be built and tested in several practical applications, including non-reciprocal multiple frequency propagation, multi-rainbow trapping, and cloaking by sound cancelation.