The physico-chemical properties of Melamine Formaldehyde (MF) based thermosets are largely influenced by the degree of polymerization (DP) in the underlying resin. On-line supervision of the turbidity point by means of vibrational spectroscopy has recently emerged as a promising technique to monitor the DP of MF resins. However, spectroscopic determination of the DP relies on chemometric models, which are usually sensitive to drifts caused by instrumental and/or sample-associated changes occurring over time. In order to detect the time point when drifts start causing prediction bias, we here explore a universal drift detector based on a faded version of the Page-Hinkley (PH) statistic, which we test in three data streams from an industrial MF resin production process. We employ committee disagreement (CD), computed as the variance of model predictions from an ensemble of partial least squares (PLS) models, as a measure for sample-wise prediction uncertainty and use the PH statistic to detect changes in this quantity. We further explore supervised and unsupervised strategies for (semi-)automatic model adaptation upon detection of a drift. For the former, manual reference measurements are requested whenever statistical thresholds on Hotelling's T2 and/or Q-Residuals are violated. Models are subsequently re-calibrated using weighted partial least squares in order to increase the influence of newer samples, which increases the flexibility when adapting to new (drifted) states. Unsupervised model adaptation is carried out exploiting the dual antecedent-consequent structure of a recently developed fuzzy systems variant of PLS termed FLEXFIS-PLS. In particular, antecedent parts are updated while maintaining the internal structure of the local linear predictors (i.e. the consequents). We found improved drift detection capability of the CD compared to Hotelling's T2 and Q-Residuals when used in combination with the proposed PH test. Furthermore, we found that active selection of samples by active learning (AL) used for subsequent model adaptation is advantageous compared to passive (random) selection in case that a drift leads to persistent prediction bias allowing more rapid adaptation at lower reference measurement rates. Fully unsupervised adaptation using FLEXFIS-PLS could improve predictive accuracy significantly for light drifts but was not able to fully compensate for prediction bias in case of significant lack of fit w.r.t. the latent variable space.
1 November 2011 in Chemometrics and Intelligent Laboratory Systems
In industrial quality control, especially in the field of manufacturing defect detection, deep learning plays an increasingly critical role. However, the efficacy of these advanced models is often hindered by their need for large-scale, annotated datasets. Moreover, these datasets are mainly based on RGB images, which are very different from X-ray images. Addressing this limitation, our research proposes a methodology that incorporates domain-specific self-supervised pretraining techniques using X-ray imaging to improve defect detection capabilities in manufacturing products. We employ two pretraining approaches, SimSiam and SimMIM, to refine feature extraction from manufacturing images. The pretraining stage is carried out using an industrial dataset of 27,901 unlabeled X-ray images from a manufacturing production line. We analyze the performance of the pretraining against transfer-learning-based methods in a complex defect detection scenario using a Faster R-CNN model. We conduct evaluations on both a proprietary industrial dataset and the publicly available GDXray dataset. The findings reveal that models pretrained with domain-specific X-ray images consistently outperform those initialized with ImageNet weights. Notably, Swin Transformer models show superior results in scenarios rich in labeled data, whereas CNN backbones are more effective in limited-data environments. Moreover, we underscore the enhanced ability of the models pretrained with X-ray images in detecting critical defects, crucial for ensuring safety in industrial settings. Our study offers substantial evidence of the benefits of self-supervised learning in manufacturing defect detection, providing a solid foundation for further research and practical applications in industrial quality control.
1 July 2022 in Revista de Psiquiatría y Salud Mental
The complexity involved in the process of real-time data-driven monitoring dynamic systems for predicted maintenance is usually huge. Up to certain extent, any data-driven approach is sensitive to data preprocessing, understood as any data treatment prior to the application of the monitoring model, being sometimes crucial for the final development of the employed monitoring technique. The aim of this work is to quantify the sensitiveness of data-driven predictive maintenance models in dynamic systems in an exhaustive way. We consider a couple of predictive maintenance scenarios, each of them defined by some public available data. For each scenario, we consider its properties and apply several techniques for each of the successive preprocessing steps, e.g., data cleaning, missing values treatment, outlier detection, feature selection, or imbalance compensation. The pretreatment configurations, i.e., sequential combinations of techniques from different preprocessing steps, are considered together with different monitoring approaches, in order to determine the relevance of data preprocessing for predictive maintenance in dynamical systems.
8 August 2024 in International Journal of Solids and Structures