Fault Location (FL) is the key to fast clearance of faults and minimizing the duration and costs of these faults. Power networks manifest a complex behavior due to their intricate topological connections, phase asymmetry and time-varying loads. Traveling Wave (TW) analysis has been proven to provide accurate estimates of fault locations on power system transmission lines. In this project, a novel selfadaptive approach for the deployment of TWs for robust fault-location in power systems has been proposed. An ensemble approach is suggested for robust detection of the Time Of Arrival (TOA) of traveling waves using a large group of wavelet types. The implemented approach was tried on real-world measurement data from a transmissionlevel substation. The robustness of the method was tested in different noise scenarios. The results show a significantly accurate and robust performance of the scheme for determining fault locations.