Condition monitoring of power electronics increases the safety and reliability of autonomous driving and electrical vehicles. Therefore, this thesis presents a concept for the detection of assembly and interconnection faults in power semiconductors during operation. Based on simulation and measurement results of a test arrangement, three detection algorithms using the readings of external temperature sensors are presented. Experimental results show that common degradation mechanisms are reliably detected and fault growth can be observed. Furthermore, two machine-learning classification approaches are presented to complement the detection algorithms.