Article
Computer Science, Artificial Intelligence
Quanqing Que, Fang Yang, Defu Zhang
Summary: This paper proposes a deep reinforcement learning model to solve the three-dimensional packing problem. The method utilizes Transformer architecture as the policy network and trains the network using Proximal Policy Optimization. Compared with previous approaches, our method presents a novel state representation and introduces plane features, achieving state-of-the-art results for solving the three-dimensional packing problem using deep reinforcement learning.
EXPERT SYSTEMS WITH APPLICATIONS
(2023)
Article
Computer Science, Interdisciplinary Applications
Bram Cals, Yingqian Zhang, Remco Dijkman, Claudy van Dorst
Summary: In this paper, the authors propose a Deep Reinforcement Learning approach combined with heuristics to optimize order picking in warehouses, showing better performance than proposed heuristics in most cases and demonstrating a different learned strategy from hand-crafted heuristics.
COMPUTERS & INDUSTRIAL ENGINEERING
(2021)
Article
Environmental Sciences
Zhiwei Jia, Haoliang Zheng, Rongjie Wang, Wenguang Zhou
Summary: In this article, a new federated learning framework is proposed to address the issues of data scarcity and data silos in aircraft feature detection. The FedDAD algorithm, which is suitable for aircraft detection in SAR images, is used to optimize the global model. The client models trained through federated cooperation have an advantage in detecting aircraft with unknown scenarios or attitudes while remaining sensitive to local datasets.
Article
Engineering, Industrial
Fatima Ezzahra Achamrah, Ali Attajer
Summary: Unlike mass production systems, reconfigurable cyber-physical systems (RCPMS) change their structures throughout missions and self-adjust production in response to demand requirements. This paper proposes a model for selective maintenance in RCPMS with imperfect repairs, integrating uncertainties from imperfect observations of components' health status. A deep reinforcement learning framework is used to solve the resulting multi-objective and combinatorial optimization problem. Decision values and the Analytical Hierarchy Process are employed for adjusting priorities and objective functions.
INTERNATIONAL JOURNAL OF PRODUCTION RESEARCH
(2023)
Article
Engineering, Civil
Jingwen Li, Liang Xin, Zhiguang Cao, Andrew Lim, Wen Song, Jie Zhang
Summary: Recent trend focuses on applying deep reinforcement learning to solve vehicle routing problem and address challenges in pairing and precedence relationships in pickup and delivery problem. Research utilizes novel neural network with heterogeneous attention mechanism to empower policy and automate node selection.
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
(2022)
Article
Computer Science, Artificial Intelligence
Rui Zhu, Faisal Aqlan, Richard Zhao, Hui Yang
Summary: This study integrates eye-tracking technology and virtual reality to evaluate and enhance the problem-solving skills of engineering students in real-world problems. By modeling the problem-solving process in manufacturing systems and analyzing the eye-tracking data with a data-driven model, the results show that the joint model is more effective than the individual model.
EXPERT SYSTEMS WITH APPLICATIONS
(2022)
Article
Engineering, Civil
Zhihan Lv, Shaobiao Zhang, Wenqun Xiu
Summary: This study aims to improve the intelligent transportation system using deep learning algorithms, leading to enhanced data transmission performance, accurate prediction, and effective path adjustment strategy to suppress congestion spread.
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
(2021)
Article
Automation & Control Systems
Miguel Neves, Pedro Neto
Summary: This paper proposes an approach to using deep reinforcement learning (DRL) in assembly sequence planning (ASP). The approach introduces parametric actions and two different reward signals, and compares the performance of different deep RL algorithms in different scenarios, while also comparing them to tabular Q-Learning. The results demonstrate the potential of deep reinforcement learning in assembly sequence planning problems with human interaction.
INTERNATIONAL JOURNAL OF ADVANCED MANUFACTURING TECHNOLOGY
(2022)
Article
Computer Science, Artificial Intelligence
Qiang Zhou, Yefei Yang, Shaochuan Fu
Summary: This study considers the impact of reference price effects on joint pricing and inventory management systems, and proposes a deep reinforcement learning algorithm to maximize the retailer's revenue. Through experiments, it is found that the retailer should not ignore the influence of current prices on future demand, and should adjust sales prices and order-up-to levels based on customers' ability to remember previous prices.
EXPERT SYSTEMS WITH APPLICATIONS
(2022)
Article
Automation & Control Systems
Zhengxuan Ling, Xinyu Tao, Yu Zhang, Xi Chen
Summary: This article explores the transformation of the Traveling Salesman Problem (TSP) to an image representation suitable for deep learning, using a Fully Convolutional Network (FCN) to learn the mapping from feasible regions to optimal solutions. The results show excellent performance and strong generalization capabilities.
IEEE TRANSACTIONS ON SYSTEMS MAN CYBERNETICS-SYSTEMS
(2021)
Article
Automation & Control Systems
Tao Feng, Jilie Zhang, Yin Tong, Huaguang Zhang
Summary: This paper addresses the consensusability problem for single-input discrete-time multi-agent systems over directed graphs using the LQR design method. It is shown that the maximum consensus region corresponds to the largest gain margin of LQR, and the necessary and sufficient condition for consensusability is derived by solving an ARE. The developed framework allows solving the consensusability problem even when the agents' models are completely unavailable.
Article
Computer Science, Artificial Intelligence
Soham Chattopadhyay, Laila Zary, Chai Quek, Dilip K. Prasad
Summary: A novel approach for motivation detection using EEG signals is proposed in this paper, which effectively addresses the issues of overfitting and vanishing gradient in small datasets through residual-in-residual architecture of convolutional neural network. The motivation state during learning can be accurately detected using alpha and beta wave signals, achieving 89% and 88% accuracy respectively.
EXPERT SYSTEMS WITH APPLICATIONS
(2021)
Article
Engineering, Multidisciplinary
Yakin Hajlaoui, Amel Jaoua, Safa Bhar Layeb
Summary: This article proposes a new deep reinforcement learning (RL) model to solve the single container loading problem. Experimental results show that the model performs well compared to existing heuristics and has good generalization capability. However, there is still an optimization gap compared to the heuristics in operations research literature. Additionally, the article discusses the benefits of training the model under different levels of variability.
ENGINEERING OPTIMIZATION
(2023)
Article
Computer Science, Interdisciplinary Applications
Juan S. Toquica, Patricia S. Oliveira, Witenberg S. R. Souza, Jose Mauricio S. T. Motta, Dibio L. Borges
Summary: This paper proposes two solutions for the inverse kinematic problem of an industrial parallel robot: a closed analytical form and a Deep Learning approximation model based on three different networks. The algorithms based on these three machine learning techniques were implemented in a tensorflow environment, analyzing the complexity of inverse kinematics and providing a novel method for solving and validating parallel robot models.
COMPUTERS & INDUSTRIAL ENGINEERING
(2021)
Article
Automation & Control Systems
Liang Xin, Wen Song, Zhiguang Cao, Jie Zhang
Summary: The article proposes a novel step-wise scheme to remove visited nodes in each node selection step, addressing the issue of suboptimal policies in routing problems. By applying this scheme, the performance of two deep models is significantly improved, and an approximate step-wise TAM model is introduced to reduce computational complexity.
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
(2021)
Article
Green & Sustainable Science & Technology
Mercedes Grijalvo Martin, Antonia Pacios Alvarez, Joaquin Ordieres-Mere, Javier Villalba-Diez, Gustavo Morales-Alonso
Summary: The industry is now in the era of the Fourth Industrial Revolution, where traditional manufacturing firms are implementing new maintenance innovations based on digitalisation and data-driven approaches. New equipment maintenance business models may require new organisational approaches at various levels, with vertical, horizontal, and transverse integration. A new prescriptive maintenance business model for equipment exploiting digitalisation opportunities is proposed, with discussions on social value and alignment with the Sustainable Development Goals.
Article
Physics, Multidisciplinary
Javier Villalba-Diez, Juan Carlos Losada, Rosa Maria Benito, Ana Gonzalez-Marcos
Summary: This study examines how the relationship between a subordinate reporting to two leaders affects the alignment of the latter with the company's strategic objectives in an Industry 4.0 environment. The research shows that when leaders communicate with each other, reporting nodes need to have an alignment probability higher than 90% to support the leader node.
Article
Physics, Multidisciplinary
Javier Villalba-Diez, Juan Carlos Losada, Rosa Maria Benito, Daniel Schmidt
Summary: This study explores the impact of the relationship between subordinates reporting to a leader on the leader's alignment with company's strategic objectives in an Industry 4.0 environment. Findings suggest that the leader's alignment probability is never higher than the average alignment value of the subordinates, recommending Industry 4.0 leaders not to add hierarchical levels without achieving high levels of stability in the lower levels.
Article
Chemistry, Analytical
Javier Villalba-Diez, Miguel Gutierrez, Mercedes Grijalvo Martin, Tomas Sterkenburgh, Juan Carlos Losada, Rosa Maria Benito
Summary: With the rise of Industry 4.0, real-time monitoring of manufacturing processes through sensor networks has become possible, leading to challenges in deterministic analysis. Bayesian decision networks and JIDOKA offer solutions to this issue, while quantum digital twins show promise in modeling complex sensor networks with high computational performance.
Article
Chemistry, Analytical
Javier Villalba-Diez, Joaquin Ordieres-Mere
Summary: The aim is to use IIoT technology and advanced data processing to promote integration strategies, aiming for a better understanding of information processing and increased human-machine integration for appropriate management. The paper evaluates how human-machine integration helps explain variability in value creation processes, through action research in different case studies.
Article
Chemistry, Multidisciplinary
Javier Villalba-Diez, Martin Molina, Daniel Schmidt
Summary: The goal of this work is to evaluate a deep learning algorithm designed for predicting the topological evolution of dynamic complex non-Euclidean graphs, and to showcase a methodology for link prediction in data generated from social media platforms like Twitter. The evaluation results indicate that the algorithm performs with high accuracy in predicting links within a retweet social network.
APPLIED SCIENCES-BASEL
(2021)
Article
Chemistry, Analytical
Antonio Sanchez-Herguedas, Angel Mena-Nieto, Francisco Rodrigo-Munoz, Javier Villalba-Diez, Joaquin Ordieres-Mere
Summary: This paper presents a methodology based on the z transform and a semi-Markovian approach to address the problems encountered when making decisions on optimal industrial preventive maintenance intervals using right-censored data. The methodology is applied to a case study of large marine engine maintenance to demonstrate its usefulness.
Article
Chemistry, Analytical
Javier Villalba-Diez, Ana Gonzalez-Marcos, Joaquin B. Ordieres-Mere
Summary: The objective of this letter is to study the optimal partitioning of value stream networks into two classes with maximized connections between them. Such problems are frequently found in system designs, such as communication network configuration and industrial applications with topological characteristics that enhance network resilience. The main focus is to improve the Max-Cut algorithm proposed in the quantum approximate optimization approach (QAOA) for a more efficient implementation. The letter also discusses related problems and suggests further research questions.
Article
Chemistry, Analytical
Marco Andres Luna, Mohammad Sadeq Ale Isaac, Ahmed Refaat Ragab, Pascual Campoy, Pablo Flores Pena, Martin Molina
Summary: This paper discusses the problems and solutions of fast coverage path planning for multiple UAVs. It proposes three methods for path assignment and verifies them through simulation and real-world experiments. The results show that the Powell optimized bin packing trajectory planner generates optimal UAV paths in minimum time.
Article
Multidisciplinary Sciences
Javier Villalba-Diez, Ana Gonzalez-Marcos, Joaquin Ordieres-Mere
Summary: This paper proposes a quantum framework for analyzing integrated systems in Industry 4.0 more efficiently. By using a novel configuration of distributed quantum circuits, the formation of industrial value chains can be evaluated. Two different mechanisms for integrating information between circuits operating at different layers are compared, enabling both linear and nonlinear behaviors while keeping the complexity bounded. The integration effects between different quantum cyber-physical digital twin models are discussed in the context of Industry 4.0 when considering component health.
SCIENTIFIC REPORTS
(2022)
Article
Chemistry, Analytical
Pablo Flores Pena, Marco Andres Luna, Mohammad Sadeq Ale Isaac, Ahmed Refaat Ragab, Khaled Elmenshawy, David Martin Gomez, Pascual Campoy, Martin Molina
Summary: This paper proposes the design of the communications, control systems, and navigation algorithms of a multi-UAV system focused on remote sensing operations. A new controller based on a compensator and a nominal controller is designed to dynamically regulate the UAVs' attitude. The navigation system addresses the multi-region coverage trajectory planning task using a new approach to solve the TSP-CPP problem. The combination of the proposed navigation techniques and control strategy was simulated to optimize the controller's parameters, and the results demonstrate the robustness of the controller and optimal performance of the route planner.
Article
Environmental Sciences
Javier Rodriguez-Vazquez, Miguel Fernandez-Cortizas, David Perez-Saura, Martin Molina, Pascual Campoy
Summary: This paper proposes a novel semi-supervised approach for accurate counting and localization of tropical plants in aerial images without labeled data. The approach utilizes deep learning and domain adaptation to handle the domain shifts between training and test data, which is a common challenge in agricultural applications. By using unsupervised domain alignment and pseudolabeling, the method adapts a model trained on a labeled source dataset to an unlabeled target dataset. Experimental results demonstrate the effectiveness of this approach in counting pineapple plants in aerial images under significant domain shifts, achieving a reduction in counting error of up to 97% (1.42 in absolute count) compared to the supervised baseline (48.6 in absolute count).
Article
Remote Sensing
Mohammad Sadeq Ale Isaac, Marco Andres Luna, Ahmed Refaat Ragab, Mohammad Mehdi Ale Eshagh Khoeini, Rupal Kalra, Pascual Campoy, Pablo Flores Pena, Martin Molina
Summary: This paper introduces a medium-scale hexacopter, called the Fan Hopper, which investigates the optimum control possibilities for a fully autonomous mission carrying a heavy payload. The research reveals that tuned Electric Ducted Fan (EDF) engines function dramatically for large payloads.