During underwater intervention missions, one of the critical issues, from the human-machine interaction perspective, concerns the operator's stress, which should be able to control the intervention system, mainly due to the high complexity of information displayed through specifically designed Graphical User Interface, and the usually available master-slave control architecture. So, this paper introduces a new approach with the aim to minimize the aforementioned drawbacks and, in the meantime, increasing the user's immersive feeling. Preliminary results are discussed highlighting the pros and cons of this novel procedure.