Tests that validate the operation of handsets into carriers' networks are very important to ensure the quality of such devices. In this way, carriers require a rigid process of certification before the delivery of these handsets into the market. Note, however, that more thorough testing prior to deployment and a reduced time to market are competing objectives that handset manufacturers must try to optimize. In this context, we have investigated ways to use the concept of autonomy, via the autonomic computing metaphor, to act in situations where a manual, or ever a simple automation control, is not enough to bring up the efficiency of our test process. Details of our autonomic model, the application domain and the results obtained so far are discussed in this paper.