<p>This work presents an intelligent method for fault identification in High Voltage Direct Current (HVDC) systems based on Source Current Converters (CSC) technology. To perform this function the proposed method uses only local signals sampled at the rectifier substation, composing a time-domain approach able to identify the faulted region in CSC-HVDC systems. The intelligent time-domain method is based on a shallow Artificial Neural Network, which allows its hardware implementation and practical evaluation by using RTDS. During the development of the proposed method the CIGRE’s reference CSC-HVDC system was used, considering different operating conditions and fault characteristics. Regardless of the operating or fault conditions, the solution presented here was always able to correctly identify the faulted region, whether in computational environment or implemented in hardware. To validate the proposed solution a second CSC-HVDC system was used and, once again, promising results in terms of accuracy and response time were observed.</p>