Dissecting a data-driven prognostic pipeline: A powertrain use case

Giordano, Danilo; Pastor, Eliana; Giobergia, Flavio; Cerquitelli, Tania; Baralis, Elena; Mellia, Marco; Neri, Alessandra; Tricarico, Davide

doi:10.1016/j.eswa.2021.115109

Cited by 11 publications

(13 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Evaluates each feature's relevance to the target variable while also seeking to minimize redundancy between the selected features [10,16].…”

Section: Wrapper Minimum Redundancy Maximum Relevancementioning

confidence: 99%

“…This broad spectrum of techniques showcases the adaptability of machine learning methods. This research has also extended to other areas, such as predicting failures of the high-pressure fuel system [10] and failure prediction in forced blowers [27]. The work [10] applied LR, RF, XGBoost, SVM, and MLP NN with feature selection using mRMR, while in [27], LR, SVM, KNN, XGBoost, and RF techniques were used without feature selection.…”

Section: Machine Learning Techniques Used To Predict Industrial Machi...mentioning

confidence: 99%

“…The literature on machine failure prediction covers a wide variety of approaches and techniques. Many studies focus on developing machine learning algorithms [9][10][11][24][25][26][27][28][29][30]. Others explore methods of processing signals acquired from machines through sensors [25,31].…”

Section: Machine Learning Techniques Used To Predict Industrial Machi...mentioning

confidence: 99%

“…Furthermore, by highlighting the most important features, it is possible to simplify the model, reducing its complexity and making it more computationally efficient [2][3][4][5]8]. This step involves choosing the most relevant or informative dataset characteristics to be used in model construction, improving model performance, and reducing dimensionality, interpretability, and computational savings, among other benefits [9][10][11][12].…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Impacts of Feature Selection on Predicting Machine Failures by Machine Learning Algorithms

Bezerra,

Oliveira Neto,

Cervi

et al. 2024

Applied Sciences

View full text Add to dashboard Cite

In the context of Industry 4.0, managing large amounts of data is essential to ensure informed decision-making in intelligent production environments. It enables, for example, predictive maintenance, which is essential for anticipating and identifying causes of failures in machines and equipment, optimizing processes, and promoting proactive management of human, financial, and material resources. However, generating accurate information for decision-making requires adopting suitable data preprocessing and analysis techniques. This study explores the identification of machine failures based on synthetic industrial data. Initially, we applied the feature selection techniques Principal Component Analysis (PCA), Minimum Redundancy Maximum Relevance (mRMR), Neighborhood Component Analysis (NCA), and Denoising Autoencoder (DAE) to the collected data and compared their results. In the sequence, a comparison among three widely known machine learning classifiers, namely Random Forest (RF), Support Vector Machine (SVM), and Multilayer Perceptron neural network (MLP), was conducted, with and without considering feature selection. The results showed that PCA and RF were superior to the other techniques, allowing the classification of failures with rates of 0.98, 0.97, and 0.98 for the accuracy, precision, and recall metrics, respectively. Thus, this work contributes by solving an industrial problem and detailing techniques to identify the most relevant variables and machine learning algorithms for predicting machine failures that negatively impact production planning. The findings provided by this study can assist industries in giving preference to employing sensors and collecting data that can contribute more effectively to machine failure predictions.

show abstract

“…Evaluates each feature's relevance to the target variable while also seeking to minimize redundancy between the selected features [10,16].…”

Section: Wrapper Minimum Redundancy Maximum Relevancementioning

confidence: 99%

Section: Machine Learning Techniques Used To Predict Industrial Machi...mentioning

confidence: 99%

Section: Machine Learning Techniques Used To Predict Industrial Machi...mentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Impacts of Feature Selection on Predicting Machine Failures by Machine Learning Algorithms

Bezerra,

Oliveira Neto,

Cervi

et al. 2024

Applied Sciences

View full text Add to dashboard Cite

show abstract

“…The performance of the proposed architecture is measured against a MLP, displaying less estimation error and faster prediction times. Giordano et al (2021) introduce a prognostic pipeline for a high-pressure fuel system. The authors evaluate the performance of SVM and MLP architectures by varying different training parameters like total number of features, training size and the hyperparameters.…”

Section: Prognosticsmentioning

confidence: 99%

The asset administration shell as enabler for predictive maintenance: a review

Rahal,

Schwarz,

Sahelices

et al. 2023

J Intell Manuf

View full text Add to dashboard Cite

The emergence of the Internet of Things and the interconnection of systems and machines enables the idea of Industry 4.0, a new industrial paradigm with a strong focus on interaction and communication between physical and digital entities, leading to the creation of cyber-physical systems. The digital twin and the standard for the Asset Administration Shell are concepts derived from Industry 4.0 that exploit the advantages of connecting the physical and virtual domains, improving the management and display of the collected data. Furthermore, the increasing availability of data has enabled the implementation of data-driven approaches, such as machine and deep learning models, for predictive maintenance in industrial and automotive applications. This paper provides a two-dimensional review of the Asset Administration Shell and data-driven methods for predictive maintenance, including fault diagnosis and prognostics. Additionally, a digital twin architecture combining the Asset Administration Shell, predictive maintenance and data-driven methods is proposed within the context of the WaVe project.

show abstract