Predictive maintenance (PdM) is one of the most powerful maintenance techniques based on the estimation of the remaining useful life (RUL) of machines. Accurately estimating the RUL is crucial to ensure the effectiveness of PdM. However, current methods have limitations in fully exploring condition monitoring data, particularly vibration signals, for RUL estimation. To address these challenges, this research presents a novel Robust Multi-Branch Deep Learning (Robust-MBDL) model. Robust-MBDL stands out by leveraging diverse data sources, including raw vibration signals, time–frequency representations, and multiple feature domains. To achieve this, it adopts a specialized three-branch architecture inspired by efficient network designs. The model seamlessly integrates information from these branches using an advanced attention-based Bi-LSTM network. Furthermore, recognizing the importance of data quality, Robust-MBDL incorporates an unsupervised LSTM-Autoencoder for noise reduction in raw vibration data. This comprehensive approach not only overcomes the limitations of existing methods but also leads to superior performance. Experimental evaluations on benchmark datasets such as XJTU-SY and PRONOSTIA showcase Robust-MBDL’s efficacy, particularly in rotating machine health prognostics. These results underscore its potential for real-world applications, heralding a new era in predictive maintenance practices.

The Industrial Internet of Things (IIoT), which integrates sensors into the manufacturing system, provides new paradigms and technologies to industry. The massive acquisition of data, in an industrial context, brings with it a number of challenges to guarantee its quality and reliability, and to ensure that the results of data analysis and modelling are accurate, reliable, and reflect the real phenomena being studied. Common problems encountered with real industrial databases are missing data, outliers, anomalies, unbalanced classes, and non-exhaustive historical data. Unlike papers present in the literature that respond to those problems in a dissociated way, the work performed in this article aims to address all these problems at once. A comprehensive framework for data flow encompassing data acquisition, preprocessing, and machine class classification is proposed. The challenges of missing data, outliers, and anomalies are addressed with critical and novel class outliers distinguished. The study also tackles unbalanced class classification and evaluates the impact of missing data on classification accuracy. Several machine learning models for the operating state classification are implemented. The study also compares the performance of the proposed framework with two existing methods: the Histogram Gradient Boosting Classifier and the Extreme Gradient Boosting classifier. It is shown that using “hard voting” ensemble learning methods to combine several classifiers makes the final classifier more robust to missing data. An application is carried out on data from a real industrial dataset. This research contributes to narrowing the theory–practice gap in leveraging IIoT technologies, offering practical insights into data analytics implementation in real industrial scenarios.