This paper proposes an automatic procedure for condition monitoring. It represents a valuable tool for the maintenance of expensive and spread systems, such as wind turbine farms. Thanks to data-driven signal processing algorithms, the proposed solution is fully automatic. The paper briefly describes all of the steps involved in the processing, from pre-processing of the acquired signals to interpretation of the generated results. It starts with an angular resampling method with speed measurement correction. Next comes a data validation step, in both time/angular and frequency/ order domains. After the pre-processing, the spectral components of the analysed signal are identified and classified in several classes, from sine wave to narrowband components. This spectral peak detection and classification allows for the extraction of the harmonic and side-band series, which may be part of the spectral content of the signal. Moreover, the detected spectral patterns are associated with the characteristic frequencies of the investigated system. Based on the detected side-band series, full-band demodulation is performed. At each step, the diagnosis features are computed and dynamically tracked, signal by signal. Finally, system health indicators are proposed to provide a conclusion on the condition of the investigated system. Altogether the abovementioned steps create a self-sufficient tool for the robust diagnosis of mechanical faults. The paper presents the performance of the proposed method on real-world signals from a wind turbine drive train.