Article
Automation & Control Systems
Alberto Jaenal, Jose-Raul Ruiz-Sarmiento, Javier Gonzalez-Jimenez
Summary: This paper presents a general deep learning architecture, MachNet, that addresses the heterogeneity of Industry 4.0-PdM solutions and is capable of handling various PdM problems. The modular architecture allows for an arbitrary number and type of sensors, and the integration of prior information. Experimental results show that MachNet achieves excellent performance in health state and remaining useful life estimation.
ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE
(2024)
Article
Green & Sustainable Science & Technology
Mustufa Haider Abidi, Muneer Khan Mohammed, Hisham Alkhalefah
Summary: With the rise of the fourth industrial revolution, the application of artificial intelligence in manufacturing is becoming more common, specifically in the importance of Predictive Maintenance (PdM). In order to improve facility lifespans, components must be repaired in advance to prevent failure.
Article
Computer Science, Artificial Intelligence
Rezvaneh Sahba, Reza Radfar, Ali Rajabzadeh Ghatari, Alireza Pour Ebrahimi
Summary: This paper proposes a novel PdM framework based on RAMI 4.0 to reduce operation and maintenance costs in the broadcasting chain, with practical outcomes and recommendations from experiments conducted using IRIB. The framework outperforms best-evaluated methods in terms of acceptance, showcasing the benefits of the proposed DSR approach.
ADVANCED ENGINEERING INFORMATICS
(2021)
Article
Computer Science, Interdisciplinary Applications
Qiushi Cao, Cecilia Zanni-Merk, Ahmed Samet, Christoph Reich, Francois de Bertrand de Beuvron, Arnold Beckmann, Cinzia Giannetti
Summary: In the context of Industry 4.0, smart factories utilize advanced technologies for production monitoring and analysis, but the heterogeneous nature of industrial data leads to complex knowledge extraction. In order to achieve predictive maintenance, symbolic AI technologies are required. KSPMI is a knowledge-based system developed based on a hybrid approach utilizing both statistical and symbolic AI technologies.
ROBOTICS AND COMPUTER-INTEGRATED MANUFACTURING
(2022)
Review
Chemistry, Multidisciplinary
Mounia Achouch, Mariya Dimitrova, Khaled Ziane, Sasan Sattarpanah Karganroudi, Rizck Dhouib, Hussein Ibrahim, Mehdi Adda
Summary: Predictive maintenance plays a key role in the fourth industrial revolution, introducing a digital version of machine maintenance to maximize efficiency, reduce downtime and costs, and improve production quality and speed.
APPLIED SCIENCES-BASEL
(2022)
Article
Computer Science, Information Systems
Yyi Kai Teoh, Sukhpal Singh Gill, Ajith Kumar Parlikad
Summary: The assets in Industry 4.0 are divided into physical, virtual, and human. The innovation and popularization of ubiquitous computing enhance the usage of smart devices for asset identification and tracking. The generated data from the Industrial Internet of Things ease information visibility and process automation. Virtual assets include the data produced from IIoT. Predictive maintenance enables businesses to decide, such as repairing or replacing the component before an actual failure that affects the whole production line. Therefore, Industry 4.0 requires effective asset management for task optimization and predictive maintenance.
IEEE INTERNET OF THINGS JOURNAL
(2023)
Article
Engineering, Industrial
Luca Pinciroli, Piero Baraldi, Enrico Zio
Summary: This paper provides a comprehensive review of maintenance optimization from different and complementary perspectives. It analyzes the knowledge, information, and data that can be used for maintenance optimization within the Industry 4.0 paradigm. The paper discusses the objectives of optimization, maintenance features to be optimized, challenges, and trends, emphasizing the need for methods that can handle heterogeneous data, uncertainties, and multiple optimization objectives.
RELIABILITY ENGINEERING & SYSTEM SAFETY
(2023)
Article
Engineering, Industrial
Camilla Lundgren, Cecilia Berlin, Anders Skoogh, Anders Kallstrom
Summary: Digitalised manufacturing provides new opportunities for running and maintaining manufacturing plants, but there is little attention given to leadership practices in this area from a socio-technical perspective. This paper presents findings from interviews with maintenance managers in the Swedish manufacturing industry, offering unique insights into the challenges faced by leaders in maintenance within the context of digitalised manufacturing. The authors propose a comprehensive consideration model for leadership that supports the development of functional maintenance organisations in the era of pervasive digitalisation.
INTERNATIONAL JOURNAL OF PRODUCTION RESEARCH
(2023)
Review
Engineering, Industrial
Foivos Psarommatis, Gokan May, Victor Azamfirei
Summary: In order to provide guidance for future research on Industry 4.0 maintenance, this study analyzed 344 eligible journal papers published between 2013 and 2022. The findings reveal key trends in advanced maintenance techniques and the consolidation of traditional maintenance concepts, driven by the increasing adoption of Industry 4.0 technologies. The study emphasizes the importance of addressing sustainability factors, human aspects, and the implementation of environmental KPIs in future research.
JOURNAL OF MANUFACTURING SYSTEMS
(2023)
Review
Computer Science, Interdisciplinary Applications
Hector Canas, Josefa Mula, Manuel Diaz-Madronero, Francisco Campuzano-Bolarin
Summary: This article focuses on the advances, advantages, limitations, requirements, and methodologies in implementing the strategic Industry 4.0 (I4.0) initiative, particularly in the field of production planning. It proposes a taxonomy of I4.0 design terms and presents models, algorithms, and components used in an I4.0 setting.
COMPUTERS & INDUSTRIAL ENGINEERING
(2021)
Review
Engineering, Industrial
Jeff Morgan, Mark Halton, Yuansong Qiao, John G. Breslin
Summary: This paper provides a fundamental research review of Reconfigurable Manufacturing Systems (RMS) and explores the state-of-the-art in distributed and decentralized machine control and machine intelligence. Key areas reviewed include RMS fundamentals, machine control technologies, and machine intelligence paradigms. The paper establishes a vision for next-generation Industry 4.0 manufacturing machines with Smart and Reconfigurable (SR*) capabilities.
JOURNAL OF MANUFACTURING SYSTEMS
(2021)
Article
Engineering, Industrial
Paul-Arthur Dreyfus, Antoine Pelissier, Foivos Psarommatis, Dimitris Kiritsis
Summary: Despite the significant investments in data-based models for the expansion of Industry 4.0, there has been insufficient effort to ensure their maintenance. This paper introduces a problem-oriented methodology to address the maintenance of industrial data-based models. The methodology includes concept-drift identification, pre-selection of solutions, optimization, and a causal concept-drift classification system. The results of the methodology show promise, but further research is required on concept drift identification and the relationship between concept-drift characteristics and drift detection.
JOURNAL OF MANUFACTURING SYSTEMS
(2022)
Article
Computer Science, Interdisciplinary Applications
Alvaro Garcia, Anibal Bregon, Miguel A. Martinez-Prieto
Summary: The recent COVID-19 outbreak has highlighted the need for manufacturing plants to adapt to unpredictable changes and ensure the continuity of industrial production. Small and Medium-sized Enterprises (SMEs) in traditional manufacturing face challenges in managing digital technologies and Industry 4.0 (I4.0) models with low adoption rates. This study presents a solution for human-machine technological integration in traditional manufacturing, enhancing digital strategies for workers and industrial systems.
COMPUTERS & INDUSTRIAL ENGINEERING
(2022)
Article
Chemistry, Analytical
Eduardo Garcia, Nicolas Montes, Javier Llopis, Antonio Lacasa
Summary: This article introduces a novel virtual sensor called mini-term for predictive maintenance. The mini-term, which represents the time it takes for a part of the machine to do its job, can serve as an indicator for machine failures and production measurement. Unlike other scientific solutions that require installing expensive sensors, the advantage of the mini-term is that it can be easily implemented in pre-installed systems. The article demonstrates the significant improvements brought by the use of the Miniterm 4.0 system in a factory setting, including increased technical availability, reduced mean time to repair, and decreased number of work orders.
Editorial Material
Engineering, Industrial
Rahul Rai, Manoj Kumar Tiwari, Dmitry Ivanov, Alexandre Dolgui
Summary: Machine learning techniques in the context of Industry 4.0 have significantly impacted the manufacturing industry by enabling smart factories, improving efficiency, and offering predictive insights, but they also face challenges such as big data management, real-time intelligence extraction, and cybersecurity concerns.
INTERNATIONAL JOURNAL OF PRODUCTION RESEARCH
(2021)
Article
Automation & Control Systems
Marco Maggipinto, Alessandro Beghi, Gian Antonio Susto
Summary: Detecting anomalies in production is crucial in manufacturing industries to meet quality goals and limit the number of defective products. Machine Learning-based approaches have been proven effective, but feature extraction is necessary for complex industrial data. Deep Learning can automatically learn useful representations of complex data. This paper proposes an anomaly detection pipeline using convolutional autoencoders for feature extraction, achieving improved performance in semiconductor manufacturing.
IEEE TRANSACTIONS ON AUTOMATION SCIENCE AND ENGINEERING
(2022)
Article
Automation & Control Systems
Davide Dalle Pezze, Chiara Masiero, Diego Tosato, Alessandro Beghi, Gian Antonio Susto
Summary: Predictive Maintenance technologies are attractive to Industrial Equipment producers for selling high added-value services and customized maintenance plans, but the costs associated with sensor measurements may hinder their development. Alarm Forecasting can serve as a low-cost alternative or helpful support to sensor-based Predictive Maintenance. A new formulation for the Alarm Forecasting problem is proposed in this work, framed as a multi-label classification task.
IEEE TRANSACTIONS ON AUTOMATION SCIENCE AND ENGINEERING
(2022)
Article
Engineering, Manufacturing
Hana T. T. Jebril, Martin Pleschberger, Gian Antonio Susto
Summary: The paper discusses the importance of data-driven fault detection and classification methods in the semiconductor manufacturing industry, proposing a feature extraction method based on deep convolutional autoencoders. The research demonstrates the superior performance of this method in the case study, while also providing open access to the data used in this work to foster research in the field.
IEEE TRANSACTIONS ON SEMICONDUCTOR MANUFACTURING
(2022)
Article
Engineering, Mechanical
Lucas C. Brito, Gian Antonio Susto, Jorge N. Brito, Marcus A. Duarte
Summary: This paper introduces a new approach for fault detection and diagnosis in rotating machinery, which includes feature extraction, fault detection, and fault diagnosis. Fault detection is achieved through vibration feature extraction and anomaly detection algorithms, while fault diagnosis is performed using the SHAP technique for interpretation of black-box models.
MECHANICAL SYSTEMS AND SIGNAL PROCESSING
(2022)
Article
Automation & Control Systems
Marco Maggipinto, Matteo Terzi, Gian Antonio Susto
Summary: This paper introduces a model called the Introspective Variational Classifier (IntroVAC), which learns interpretable latent subspaces by exploiting information from an additional label and improves image quality through an adversarial training strategy. The results show that IntroVAC is able to learn meaningful directions in the latent space, enabling fine-grained manipulation of image attributes.
ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE
(2022)
Editorial Material
Engineering, Manufacturing
Gian Antonio Susto, Alain Diebold, Andreas Kyek, Chia-Yen Lee, Nital S. Patel
IEEE TRANSACTIONS ON SEMICONDUCTOR MANUFACTURING
(2022)
Article
Computer Science, Information Systems
Alberto Purpura, Gianmaria Silvello, Gian Antonio Susto
Summary: This paper investigates how to train LETOR models with relevance judgments distributions and proposes five new probabilistic loss functions. The results show that LETOR models trained with relevance judgments distributions can improve performance and outperform strong baselines on multiple test collections.
JOURNAL OF THE ASSOCIATION FOR INFORMATION SCIENCE AND TECHNOLOGY
(2022)
Article
Automation & Control Systems
Francesco Simmini, Mirco Rampazzo, Fabio Peterle, Gian Antonio Susto, Alessandro Beghi
Summary: This study investigates the detection of faults in HVAC chiller systems using a data-driven approach, with KPCA capturing normal operative conditions. Test results show that KPCA outperforms linear PCA in fault detection, demonstrating good effectiveness.
IEEE TRANSACTIONS ON CONTROL SYSTEMS TECHNOLOGY
(2022)
Article
Computer Science, Artificial Intelligence
Alessandro Fabris, Stefano Messina, Gianmaria Silvello, Gian Antonio Susto
Summary: This research addresses the data documentation debt in the algorithmic fairness community by surveying datasets used in fairness research, providing standardized and searchable documentation. It also documents the most popular fairness datasets and provides in-depth analysis and alternative options. It further analyzes the datasets from the perspective of five important data curation topics and proposes best practices for curating novel resources.
DATA MINING AND KNOWLEDGE DISCOVERY
(2022)
Article
Computer Science, Artificial Intelligence
David Dandolo, Chiara Masiero, Mattia Carletti, Davide Dalle Pezze, Gian Antonio Susto
Summary: This paper proposes a fast interpretability approach called Accelerated Model-agnostic Explanations (AcME) for human-in-the-loop Machine Learning applications. AcME provides feature importance scores at both the global and local level and offers a what-if analysis tool to examine the impact of feature changes on model predictions. The approach achieves comparable explanation quality to state-of-the-art methods while greatly reducing computational time.
EXPERT SYSTEMS WITH APPLICATIONS
(2023)
Article
Computer Science, Information Systems
Tommaso Barbariol, Gian Antonio Susto
Summary: Unsupervised anomaly detection is a widely used method for finding anomalies in datasets without labels. Isolation Forest is a popular algorithm that defines an anomaly score using special trees. However, the standard algorithm needs improvement in terms of memory requirements, latency, and performances. Additionally, current anomaly detection methods do not fully utilize weak supervisions.
INFORMATION SCIENCES
(2022)
Article
Plant Sciences
Qiuran Wang, Tommaso Barbariol, Gian Antonio Susto, Bianca Bonato, Silvia Guerra, Umberto Castiello
Summary: Climbing plants require external support for vertical growth and enhanced light acquisition. Machine learning methods can accurately capture the differences in circumnutation patterns related to the presence/absence of support. Distinctive kinematic features at the junction underneath the tendrils can indicate the presence/absence of support by the plant.
Proceedings Paper
Computer Science, Artificial Intelligence
Alessandro Fabris, Stefano Messina, Gianmaria Silvello, Gian Antonio Susto
Summary: A community of researchers has conducted a survey and standardized documentation for over two hundred datasets used in algorithmic fairness research, aiming to tackle the collective data documentation debt. They have also identified the most popular fairness datasets and provided in-depth documentation for them, as well as documented alternatives to address their limitations. The assembled and summarized information on hundreds of datasets is made available to the community.
ACM CONFERENCE ON EQUITY AND ACCESS IN ALGORITHMS, MECHANISMS, AND OPTIMIZATION, EAAMO 2022
(2022)
Proceedings Paper
Computer Science, Theory & Methods
Elisa Marcelli, Tommaso Barbariol, Vincenzo Savarino, Alessandro Beghi, Gian Antonio Susto
Summary: Anomaly detection is a key process for identifying unusual data with different behavior compared to the rest of the dataset. With the increase in the amount of data to be analyzed, traditional anomaly detectors face challenges. This paper proposes a reinterpretation of the classical method, specifically designed for datasets with a large number of data points, and achieves promising results.
2022 23RD IEEE LATIN-AMERICAN TEST SYMPOSIUM (LATS 2022)
(2022)
Article
Computer Science, Interdisciplinary Applications
Alberto Purpura, Giuseppe Sartori, Dora Giorgianni, Graziella Orru, Gian Antonio Susto
Summary: Deception or faking is a significant concern in data collection through questionnaires. Previous studies have shown that individuals tend to provide fake answers when they have an advantage, such as during job application tests. Existing methods can identify overall faking attitudes but fail to detect faking patterns and specific affected responses. In this research, a self-attention-based autoencoder (SABA) model is proposed to identify faked responses solely based on a set of honest answers that may not be directly related to the final use case. The model outperforms three competitive baselines in terms of F1 score, namely an autoencoder based on feedforward layers, a distribution model, and a k-nearest-neighbor-based model.
Article
Automation & Control Systems
Subhashis Nandy
Summary: This research focuses on the design and stability analysis of nonlinear controllers for an electrically driven marine cycloidal propeller, along with estimating various parameters using the Extended Kalman Filter. The controller is defined using an efficient physics-based model and is able to accurately process multiple control signals. The robustness of the controller is assessed using Monte Carlo simulation, and its performance is evaluated through validation investigations.
CONTROL ENGINEERING PRACTICE
(2024)
Article
Automation & Control Systems
Lucas C. Borin, Guilherme Hollweg, Caio R. D. Osorio, Fernanda M. Carnielutti, Ricardo C. L. F. Oliveira, Vinicius F. Montagner
Summary: This work presents a new automated test-driven design procedure for robust and optimized current controllers applied to LCL-filtered grid-tied inverters. The design of control gains is guided by high-fidelity simulations and particle swarm optimization algorithm, considering various normal and abnormal operating conditions. The proposed design ensures superior performance compared with other current control designs.
CONTROL ENGINEERING PRACTICE
(2024)
Article
Automation & Control Systems
Wei He, Xiang Wang, Mohammad Masoud Namazi, Wangping Zhou, Josep M. Guerrero
Summary: The main objective of this paper is to develop a reduced-order adaptive state observer for a large class of DC-DC converters with constant power load, in order to estimate their unavailable states and unknown parameter and achieve an output feedback control scheme. The observer is designed using a generalized parameter estimation based observer technique and dynamic regressor extension and mixing method. The comparison study shows that the observer has the advantage of verifying the observability of the systems for exponential convergence without any extra excitation condition.
CONTROL ENGINEERING PRACTICE
(2024)
Article
Automation & Control Systems
Te Zhang, Bo Zhu, Lei Zhang, Qingrui Zhang, Tianjiang Hu
Summary: This paper introduces a control technique called time-varying uncertainty and disturbance estimator (TV-UDE) which extends the classic UDE approach to handle more complicated issues. By combining TV-UDE with a nominal dynamic output-feedback controller, robust control for uncertain second-order attitude control systems without velocity measurements is achieved. Numerical simulations and physical experiments on a 2-DOF AERO attitude helicopter platform demonstrate the effectiveness of the proposed design in reducing steady-state errors and avoiding issues caused by high-gain estimation.
CONTROL ENGINEERING PRACTICE
(2024)
Article
Automation & Control Systems
Kanishke Gamagedara, Taeyoung Lee, Murray Snyder
Summary: This paper presents the developments of flight hardware and software for a multirotor unmanned aerial vehicle capable of autonomously taking off and landing on a moving vessel in ocean environments. The flight hardware consists of a general-purpose computing module connected to a low-cost inertial measurement unit, real-time kinematics GPS, motor speed controller, and a camera through a custom-made printed circuit board. The flight software is developed in C++ with multi-threading to execute control, estimation, and communication tasks simultaneously. The proposed flight system is verified through autonomous flight experiments on a research vessel in Chesapeake Bay, utilizing real-time kinematics GPS for relative positioning and vision-based autonomous flight for shipboard launch and landing.
CONTROL ENGINEERING PRACTICE
(2024)
Article
Automation & Control Systems
Yun Zhu, Kangkang Zhang, Yucai Zhu, Pengfei Jiang, Jinming Zhou
Summary: In this study, a three-term Dynamic Matrix Control (DMC) algorithm using quadratic programming is developed and compared with the traditional two-term DMC algorithm. Simulation studies and real-life tests show that the three-term DMC algorithm outperforms the two-term DMC algorithm in control effectiveness.
CONTROL ENGINEERING PRACTICE
(2024)
Article
Automation & Control Systems
Jayu Kim, Taehoon Lee, Cheol-Joong Kim, Kyongsu Yi
Summary: This paper presents a data-based model predictive control method for a semi-active suspension system. The method utilizes a continuous damping controller and a stiffness controller to improve ride comfort and reduce vehicle pitch motion. Gaussian process regression is also used to compensate for model parameter uncertainties. The algorithm has been verified through computer simulations and vehicle tests, demonstrating its effectiveness and robustness.
CONTROL ENGINEERING PRACTICE
(2024)
Article
Automation & Control Systems
Kunpeng Zhang, Jikang Gao, Zongqi Xu, Hui Yang, Ming Jiang, Rui Liu
Summary: A improved dynamic programming model is proposed in this paper for joint operation optimization of virtual coupling of heavy-haul trains. By simultaneously optimizing the headway and energy savings, as well as performing locomotive engineering advisory analysis, significant improvements in train performance can be achieved.
CONTROL ENGINEERING PRACTICE
(2024)
Article
Automation & Control Systems
Demian Garcia-Violini, Yerai Pena-Sanchez, Nicolas Faedo, Fernando Bianchi, John V. Ringwood
Summary: This study presents a model invalidation methodology for wave energy converters (WECs) that can effectively handle dynamic uncertainty and external noise. The results indicate that neglecting dynamic uncertainty can lead to overestimation of performance, highlighting the importance of accurate dynamic description for estimating control performance.
CONTROL ENGINEERING PRACTICE
(2024)
Article
Automation & Control Systems
Shengyang Lu, Yue Jiang, Xiaojun Xu, Hanxiang Qian, Weijie Zhang
Summary: This paper proposes an adaptive heading tracking control strategy based on wheelbase changes for unmanned ground vehicles (UGVs) with variable configuration. The strategy adjusts the wheelbase according to different working conditions to optimize driving performance. The impact of changing wheelbase on sideslip angle and heading angle is analyzed, and a robust-active disturbance rejection control method is developed to achieve desired front-wheel steering angle. A torque distribution method based on tire load rate and real-time load is applied to enhance longitudinal stability.
CONTROL ENGINEERING PRACTICE
(2024)
Article
Automation & Control Systems
Domenico Dona, Basilio Lenzo, Paolo Boscariol, Giulio Rosati
Summary: This paper proposes a new method for designing minimum energy trajectories for servo-actuated systems and demonstrates its accuracy and effectiveness through numerical comparisons and experimental validation.
CONTROL ENGINEERING PRACTICE
(2024)
Article
Automation & Control Systems
Haolin Wang, Luyao Zhang, Yao Mao, Qiliang Bao
Summary: This paper proposes a method of transforming the core element of ADRC, ESO, into a novel fuzzy self-tuning observer structure to improve the stability of LOS in the electro-optical tracking system. It effectively solves the conflict between disturbance rejection ability and noise attenuation ability in traditional ESO.
CONTROL ENGINEERING PRACTICE
(2024)
Article
Automation & Control Systems
Alejandro Toro-Ossaba, Juan C. Tejada, Santiago Rua, Juan David Nunez, Alejandro Pena
Summary: This work presents the development of a myoelectric Model Reference Adaptive Controller (MRAC) with an Adaptive Kalman Filter for controlling a cable driven soft elbow exoskeleton. The proposed MRAC controller is effective in both passive and active control modes, showing good adaptability and control capabilities.
CONTROL ENGINEERING PRACTICE
(2024)
Article
Automation & Control Systems
Mehrad Jaloli, Marzia Cescon
Summary: This study presents an advanced multi-agent reinforcement learning (RL) strategy for personalized glucose regulation, which is shown to improve glucose regulation and reduce the risk of severe hyperglycemia compared to traditional therapy.
CONTROL ENGINEERING PRACTICE
(2024)
Article
Automation & Control Systems
Yingming Tian, Kenan Du, Jianfeng Qu, Li Feng, Yi Chai
Summary: This paper investigates the control strategy for PMSM with position sensor fault in railway. A learning observer-based control strategy is proposed, which achieves high-precision estimation of electromotive force and accelerates speed response.
CONTROL ENGINEERING PRACTICE
(2024)