Compound fault diagnosis in multi-joint industrial robots is a critical yet underexplored problem in industrial internet of things, where the simultaneous degradation of multiple joints poses a severe challenge for reliable operation. Unlike conventional methods limited to single-fault scenarios, this paper addresses the compositional generalization challenge—requiring models trained only on simple faults to accurately recognize unseen higher-order fault compositions. To this end, we propose StateMix Network (SMNet), a multi-stage architecture that preserves atomic joint-level representations before compositional diagnosis. Specifically, a Single-Joint Feature Extraction (SJFE) backbone extracts clean joint-private features, which are then fused by an Attention-Guided Dilated Fusion (AGDF) neck employing parallel Cascaded Dilated Convolution Blocks (CDCBs) bracketed by a dual-path attention mechanism for scale- and context-aware integration. Finally, a Mamba-based sequence mixer models long-range cross-joint dependencies to capture global fault dynamics. Extensive experiments on in-situ vibration data from a single six-joint industrial robot platform, under a strict train-on-simple/evaluate-on-complex protocol, demonstrate that SMNet consistently outperforms representative baselines in macro-Precision, Recall, and F1-score, particularly on unseen triple- and quadruple-joint compositions. Ablation and sensitivity analyses further validate the effectiveness of each module. This work presents a diagnostic approach that effectively generalizes from simple to complex fault scenarios in industrial robots.

Safe operation is a key concern for industrial robots. However, due to hardware failures and unstable data transmission issues, the multivariate time-series data generated by these axes often contain missing or corrupted signals, which severely hinders downstream tasks such as fault diagnosis. Additionally, the substantial volume of industrial data demands considerable time for training time-series imputation models and subsequent classification models. To address these challenges, this study proposes a multitask approach that serves both the data imputation and fault diagnosis tasks for industrial robots. Specifically, the parameters trained in the imputation model can be transferred to the fault diagnosis model, enhancing its performance and efficiency. A multitask method named Bidirectional-Graph Attention Networks Parallel Encoder (Bi-GATPE) is proposed, which employs a bidirectional graph attention network to capture the spatial dependencies among the various variables of industrial robots. Subsequently, a parallel encoder with Diagonal-Filter Attention is designed to model temporal correlations. This dual approach improves the accuracy and training speed for both the imputation and fault diagnosis tasks. Experimental studies based on real industrial robot datasets demonstrate that by modifying the feature fusion layer of the imputation task and sharing the trained parameters with the fault diagnosis task, the proposed method significantly accelerates the convergence of the fault diagnosis model while also improving diagnostic accuracy. The experiments also indicate that our method shows merits in the imputation and fault diagnosis tasks. The source code of Bi-GATPE is available at:https://github.com/miten073/Bi-GATPE.

The demand for disruption-free fault diagnosis of mechanical equipment under a constantly changing operation environment poses a great challenge to the deployment of data-driven diagnosis models in practice. Extant continual learning-based diagnosis models suffer from consuming a large number of labeled samples to be trained for adapting to new diagnostic tasks and failing to account for the diagnosis of heterogeneous fault types across different machines. In this paper, we use a representative mechanical equipment -rotating machinery – as an example and develop an uncertainty-aware continual learning framework (UACLF) to provide a unified interface for fault diagnosis of rotating machinery under various dynamic scenarios: class continual scenario, domain continual scenario, and both. The proposed UACLF takes a three-step to tackle fault diagnosis of rotating machinery with homogeneous-heterogeneous faults under dynamic environments. In the first step, an inter-class classification loss function and an intra-class discrimination loss function are devised to extract informative feature representations from the raw vibration signal for fault classification. Next, an uncertainty-aware pseudo labeling mechanism is developed to select unlabeled fault samples that we are able to assign pseudo labels confidently, thus expanding the training samples for faults arising in the new environment. Thirdly, an adaptive prototypical feedback mechanism is used to enhance the decision boundary of fault classification and diminish the model misclassification rate. Experimental results on three datasets suggest that the proposed UACLF outperforms several alternatives in the literature on fault diagnosis of rotating machinery across various working conditions and different machines.Note to Practitioners—This paper presents a continual fault diagnosis methodology for mechanical equipment under various working conditions across different machines with homogeneous-heterogeneous faults. On the application side, the proposed UACLF can be applied to facilitate diagnosis across a broad range of complex industrial equipment, including aerospace, automobile transmission, and wind turbines, among others. With the uncertainty-aware pseudo labeling, the proposed framework is empowered to select the samples in the new phase that we are able to reliably assign their labels. Hence, it can effectively improve mechanical equipment fault classification accuracy in the case that only a small portion of labeled fault samples is available. When training the model, given the model architecture, fault samples collected from multiple accelerometers are fed into the developed model. Four different loss functions, supervision loss, inter-class classification loss, intra-class discrimination loss, and uncertainty estimation loss, are employed to train the diagnostic model. Experiments conducted on three different laboratory datasets have demonstrated the effectiveness of the proposed framework, but have not been tested in the practical industrial applications. We will consider testing the proposed UACLF in an actual plant in future research.

In mechanical fault diagnosis (MFD), intelligent fault diagnosis (IFD) methods perform excellently regarding diagnosis accuracy. However, those methods are generally constructed with an excessive number of unprincipled parameters, resulting in uninterpretable architecture, ambiguity in the diagnosis process, and unclear decision-making basis. Thus, a traceable algorithm unrolling (TAU) network for interpretable MFD is proposed to overcome the above limitations. First, a mechanism-driven feature extractor (FE) is constructed by unrolling the iterative algorithm of sparse coding, aiming at encoding interpretable features from vibration signals. Second, a theory-based feature clustering (FC) algorithm is executed through the dynamic routing mechanism of the capsule network (CN), where the inner product serves as a measure for the association between input and output features. Finally, a post hoc interpretability strategy based on the coupling matrix is introduced to investigate how the TAU generates diagnostic results from the learned features and to verify whether these features are associated with faults, thereby enhancing the credibility of the diagnosis results. In addition, the simulation and experiment are designed to verify the decision-making mechanism and diagnostic performance of TAU. The results demonstrate that TAU makes diagnostic decisions based on the high-dimensional feature mapping associated with fault characteristic frequencies. Meanwhile, TAU outperforms the compared methods in fault diagnosis performance.

Due to high-transmission efficiency, gearboxes have become indispensable components of industrial mechanical equipment. It is paramount for gearbox fault diagnosis to extract discriminant features under strong interferences of industrial scene. However, the extraction performance of current methods is not satisfactory in interpretability and robustness. In this article, an interpretable approach named mechanism-informed neural network (MINN) is proposed for robust impulsive fault feature (IFF) extraction. First, standard auto-encoder is modified based on sparse representation to construct an unsupervised MINN. Second, a mechanism-informed dictionary is designed and embedded into MINN, which brings physical interpretability for the IFF extraction. Third, a two-stage IFF extraction framework is formulated, in which the network parameters are adaptively updated with the proposed joint optimization algorithm to achieve robust IFF extraction. Finally, comparative studies in simulation and experiment are conducted. The results demonstrate that MINN performs better in IFF extraction under strong harmonic interferences. Moreover, the extracted IFF of MINN has been analyzed and interpreted from the view of impulsive fault mechanism, which enhances the reliability.

Impulsive feature (IF) response is an essential indicator for rolling bearing fault. However, it is overwhelmed by strong noise and difficult to extract in real scenes. Although deep learning-based methods are powerful in feature extraction, their logic and extracting principles possess weak interpretability and credibility. Their further implementation is hampered. In this article, a natural modal sketching network (NMSNet) is constructed to achieve robust and credible bearing IF extraction. First, the modal response is designed as a convolutional kernel of NMSNet, and the forward propagation logic is interpreted as natural modal sketching, including modal response recovery and weighted superposition. The logic derives from the fault mechanism and brings solid credibility to NMSNet. Second, a novel correction algorithm is developed to interpret the extraction principle of NMSNet in theory and achieve noise elimination due to its filter nature. Third, NMSNet realizes adaptive modal sketching via the formulated weighted fusion strategy and training constraint. Finally, simulation and experiment have been carried out to verify the effectiveness and noise robustness of NMSNet. The fault-related interpretability analysis confirms the knowledge acquisition of NMSNet, which strengthens the credibility of IF extraction.

Rotatory machinery commonly operates in complex environments with strong noise and variable working conditions. Time-frequency representation offers a valuable method for capturing and analyzing nonstationary characteristics, making it particularly suitable for identifying transient fault-related features. However, despite these advantages, extracting robust and interpretable fault features in machinery operating under variable speeds remains a challenge with existing techniques. In this article, a novel sparse time-frequency representation (STFR) method, named matching pursuit network (MPNet) is proposed for mechanical fault diagnosis. First, a deep network structure with signal decomposition capability is constructed by well-defined interpretable matching pursuit (MP) units to automatically learn discriminative features from time-frequency inputs. Then, the weights of each effective component signal to reconstruct the raw input are designed to measure their contributions. Accordingly, the optimization criterion with structural similarity metric is produced to realize the model parameter update in an end-to-end manner. Finally, phenomenological model-based fault simulation signals and real fault signals from gearbox experiments are used for model training and testing, respectively. The results show that the proposed approach can well extract robust and interpretable time-frequency features and obviously outperforms the state-of-the-art time-frequency representation methods.

The development of Internet of Things technology provides abundant data resources for prognostics health management of industrial machinery, and data-driven methods have shown their powerful ability in the field of fault diagnosis. However, these methods have several limitations: 1) Using less labeled data to obtain higher accuracy is a challenging task, which limits the application of diagnostic models in practical applications. 2) Physics-informed knowledge is largely ignored during the modeling process, which contains a wealth of information that can reflect the harmonic drive’s health status. To address these challenges, a self-supervised fault diagnosis framework is developed by integrating prior knowledge with deep learning to improve the accuracy and reliability of diagnosis models in industrial applications. Specifically, the physics-based knowledge including 32-dimensional time domain, frequency domain, and time-frequency domain features, is first designed to provide fault information and significantly reduce the amount of data required for deep learning. Furthermore, a self-supervised knowledge embedded auto-encoder network is built by employing the prior knowledge in the multi-scale convolutional auto-encoder. With the ability to integrate prior knowledge and the self-supervised learning mechanism, the proposed method can provide a strong tool for knowledge representation and an effective solution for fault diagnosis under a few-shot industrial scenario. The experimental results conducted on a real harmonic drive fault dataset prove that the proposed network framework provides effective insights on fault diagnosis and has excellent generalizability in practical industrial applications.

In real industries, there often exist application scenarios where the target domain holds fault categories never observed in the source domain, which is an open-set domain adaptation (DA) diagnosis issue. Existing DA diagnosis methods under the assumption of sharing identical label space across domains fail to work. What is more, labeled samples can be collected from different sources, where multisource information fusion is rarely considered. To handle this issue, a multisource open-set DA diagnosis approach is developed. Specifically, multisource domain data of different operation conditions sharing partial classes are adopted to take advantage of fault information. Then, an open-set DA network is constructed to mitigate the domain gap across domains. Finally, a weighting learning strategy is introduced to adaptively weigh the importance on feature distribution alignment between known class and unknown class samples. Extensive experiments suggest that the proposed approach can substantially boost the performance of open-set diagnosis issues and outperform existing diagnosis approaches.

Recently, domain adaptation techniques have achieved great attention in solving domain-shift problems of mechanical fault diagnosis. However, existing methods mostly work under assumption that source domain and target domain share identical label spaces, which fail to handle those issues, where a large set of source data classes are available and target data only cover a subset of classes. To address this problem, a novel weighted adversarial transfer network (WATN) is proposed for partial domain fault diagnosis, in this article. Adversarial training is introduced to learn both class discriminative and domain invariant features, and a weighting learning strategy is adopted to weigh their contributions to both source classifier and domain discriminator. As such, the irrelevant source examples can be identified and filtered out, and the distribution discrepancy of shared classes between domains can be reduced. Experiments on two diagnosis data sets demonstrate that the proposed WATN achieves satisfactory performance and outperforms state-of-the-art methods.

Deep neural networks present very competitive results in mechanical fault diagnosis. However, training deep models require high computing power while the performance of deep architectures in extracting discriminative features for decision making often suffers from the lack of sufficient training data. In this paper, a transferable convolutional neural network (CNN) is proposed to improve the learning of target tasks. First, a one-dimensional CNN is constructed and pretrained based on large source task datasets. Then a transfer learning strategy is adopted to train a deep model on target tasks by reusing the pretrained network. Thus, the proposed method not only utilizes the learning power of deep network but also leverages the prior knowledge from the source task. Four case studies are considered and the effects of transfer layers and training sample size on classification effectiveness are investigated. Results show that the proposed method exhibits better performance compared with other algorithms.