Computers & Industrial Engineering最新文献

{"title":"Data-driven analysis on inventory problem for anticipatory shipping","authors":"Xinxin Ren ,&nbsp;Kuan Zeng","doi":"10.1016/j.cie.2025.111038","DOIUrl":"10.1016/j.cie.2025.111038","url":null,"abstract":"<div><div>Under the anticipatory shipping (AS) mode, online retailers predict inventory and ship items to various hubs before orders arrive. In fact, the inventory decision for each hub is essence for AS, since overage increases shipping costs, while underage incurs order losses, similar to a Newsvendor mode. In this study, we introduce a “Machine Learning - Newsvendor” framework and adopt big data analytics to optimize the AS inventory. Specifically, we integrate forecasting algorithm into optimization model and develop two algorithms, i.e., <em>XGBoost-NV</em> (eXtreme Gradient Boosting - Newsvendor) and <em>LGBM-NV</em> (Light Gradient Boosting Machine - Newsvendor), and then propose a two-stage hybrid algorithm <em>C-DT-NV</em> (Clustering-Decision Tree-Newsvendor), facilitating algorithm selection between <em>XGBoost-NV</em> and <em>LGBM-NV</em>, for homogeneous products. Through employing the algorithms over enormous amount of data in a large bakery chain, we find that <em>XGBoost-NV</em> (<em>LGBM-NV</em>) outperforms XGBoost (LGBM) in the AS inventory optimization, where the average cost decreases by 23.91% (23.57%), and <em>C-DT-NV</em> decreases average cost and inventory error further. Finally, we examine three influential factors (i.e., price, demand volatility and return quantity) in AS inventory within the “Machine Learning - Newsvendor” framework and find that, as product demand declines, the AS inventory decreases with price increasingly, while the AS inventory decreases with demand volatility significantly, regardless of the demand. In addition, as return quantity decreases, the AS inventory is more likely to increase when the demand stays high, but keeps constant when the demand is low enough.</div></div>","PeriodicalId":55220,"journal":{"name":"Computers & Industrial Engineering","volume":"203 ","pages":"Article 111038"},"PeriodicalIF":6.7,"publicationDate":"2025-03-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143654736","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Anomaly detection in manufacturing systems with temporal networks and unsupervised machine learning","authors":"Giulio Mattera ,&nbsp;Raffaele Mattera ,&nbsp;Silvestro Vespoli ,&nbsp;Emma Salatiello","doi":"10.1016/j.cie.2025.111023","DOIUrl":"10.1016/j.cie.2025.111023","url":null,"abstract":"<div><div>Traditional manufacturing systems face significant challenges in detecting operational anomalies due to the absence of advanced sensor networks and intelligent machinery commonly associated with Industry 4.0. Existing solutions often rely on sophisticated, interconnected infrastructures, which are not feasible in conventional settings. This paper introduces a novel methodology for anomaly detection tailored specifically for traditional manufacturing environments, addressing the gap in cost-effective monitoring solutions. The proposed approach models manufacturing systems as complex temporal networks, where each machine or process is represented as a node and job flows between machines form the network edges over time. The novelty of this method lies in the combination of dynamic network theory with unsupervised machine learning. Statistical features extracted from the temporal networks are processed through dimensionality reduction techniques, specifically Principal Component Analysis (PCA) and Deep Neural Autoencoders, to reduce feature complexity while preserving essential information. The reduced feature sets are then analysed using multiple unsupervised anomaly detection algorithms, including Isolation Forest, One-Class Support Vector Machine (OC-SVM), and Local Outlier Factor (LOF). This approach does not require significant infrastructure upgrades, making it suitable for traditional manufacturing plants while still aligning with Industry 4.0 paradigms. By using only normal job flow data, it provides a cost-effective solution where anomalous data is scarce. The results demonstrate that Local Outlier Factor and Isolation Forest, when combined with Autoencoder-based feature reduction, achieved an F1-score exceeding 84%, with precision close to 99% and recall at 74%. This strong performance underscores the methodology’s potential for real-world manufacturing environments, bridging the gap between traditional settings and modern Industry 4.0 paradigms.</div></div>","PeriodicalId":55220,"journal":{"name":"Computers & Industrial Engineering","volume":"203 ","pages":"Article 111023"},"PeriodicalIF":6.7,"publicationDate":"2025-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143642037","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Linking social media data and patents via Wikipedia for social problem-solving R&D","authors":"Seunghyun Lee ,&nbsp;Jiho Lee ,&nbsp;Jae-Min Lee ,&nbsp;Hong-Woo Chun ,&nbsp;Janghyeok Yoon","doi":"10.1016/j.cie.2025.111039","DOIUrl":"10.1016/j.cie.2025.111039","url":null,"abstract":"<div><div>As modern society develops and changes rapidly, social issues occur in various fields and their impact grows and evolves into social problems. Despite the potential for technology to be a solution to social issues, prior studies have focused on social issue detection, but rarely on identifying practical solutions corresponding to the detected social issues. In this study, we propose an approach that relates social media data to patents to identify existing R&amp;D solutions applicable to social issues. The approach involves 1) extracting core keywords of a detected social issue from social media; 2) expanding the core keywords to R&amp;D keywords using Wikipedia as a bridge database between the social issue and patents; 3) identifying R&amp;D solutions utilizing a search query defined with the core keywords and R&amp;D keywords of the social issue; and 4) analyzing detailed technologies of the R&amp;D solutions. This study contributes to the existing literature by proposing a new approach for linking heterogeneous datasets using Wikipedia. Additionally, this study, which identifies R&amp;D solutions to social issues, offers a new direction for social problem-solving R&amp;D as an extension of prior studies on social issues.</div></div>","PeriodicalId":55220,"journal":{"name":"Computers & Industrial Engineering","volume":"203 ","pages":"Article 111039"},"PeriodicalIF":6.7,"publicationDate":"2025-03-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143654737","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Enhancing strategic investment in construction engineering projects: A novel graph attention network decision-support model","authors":"Fatemeh Mostofi ,&nbsp;Ümit Bahadır ,&nbsp;Onur Behzat Tokdemir ,&nbsp;Vedat Toğan ,&nbsp;Victor Yepes","doi":"10.1016/j.cie.2025.111033","DOIUrl":"10.1016/j.cie.2025.111033","url":null,"abstract":"<div><div>Selecting the right investment projects is a pivotal decision-making process that can steer a company’s financial and operational future. Existing methods often fall short in merging machine learning with network-based multi-criteria decision-making (MCDM) strategies. This research presents a first-time investment network framework fed into a graph attention network (GAT) to forecast the success of construction engineering projects by leveraging their interrelated data across various decision-making parameters. Expert judgment was initially employed to filter over 33,000 investment projects based on organizational goals, project risk, and business development ratings. The refined dataset was organized into three specialized MCDM investment-decision networks: regional-based, country-level, and funding-mode-based. These networks were subsequently fed into GAT models to classify investment values. The regional-based network achieved over 99 % accuracy, the country-level and funding-mode-based networks delivered over 98 % accuracy. These insights demonstrate that while all three models maintain high accuracy, the slight variances in their performance reflect the importance of tailoring decision-support tools to specific geographical contexts. The understanding of different network structures can provide strategic decision-making insight for large-scale infrastructure investments, where even minor misclassifications can lead to substantial financial consequences.</div></div>","PeriodicalId":55220,"journal":{"name":"Computers & Industrial Engineering","volume":"203 ","pages":"Article 111033"},"PeriodicalIF":6.7,"publicationDate":"2025-03-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143631735","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"An unsupervised subdomain adaptation of cross-domain remaining useful life prediction for sensor-equipped equipments","authors":"Jianhai Yan ,&nbsp;Zhi-Sheng Ye ,&nbsp;Shuguang He ,&nbsp;Zhen He","doi":"10.1016/j.cie.2025.110967","DOIUrl":"10.1016/j.cie.2025.110967","url":null,"abstract":"<div><div>Most existing remaining useful life (RUL) prediction models assume a similarity between the data distributions in the source and target domains, but this assumption is often challenging in practical applications. Although traditional transfer learning methods could relax this assumption by minimizing the distributional differences between domains, they often ignore the fine-grained features exhibited by equipment at different health state and the inherent properties of the domains. Additionally, most models ignore the importance of the location information of the sensors equipped on the equipment, and the real-world problem of the high cost of fully labeling the data. Addressing these challenges, we introduce a novel model of unsupervised subdomain adaptation and feature disentanglement for equipment RUL prediction. The model contains four components: a feature extractor, a feature similarity component, a domain adaptation component, and a multi-task module. Specifically, the feature extractor designs a dynamic fusion network to divide the domain’s private and shared features. The feature similarity component employs a cross-attention mechanism to constrain the extracted features. The domain adaptation enables the proposed model to extract deeper subdomain-shared features. The multi-task module simultaneously divides subdomains and performs RUL prediction. The proposed model is validated using multiple tasks and compared with existing works.</div></div>","PeriodicalId":55220,"journal":{"name":"Computers & Industrial Engineering","volume":"203 ","pages":"Article 110967"},"PeriodicalIF":6.7,"publicationDate":"2025-03-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143631830","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Constraint first, shrinking next: A hybrid photovoltaic generation forecasting framework based on ensemble learning and multi-strategy improved optimizer","authors":"Jionghao Zhu ,&nbsp;Jie Liu ,&nbsp;Xiaoying Tang","doi":"10.1016/j.cie.2025.111022","DOIUrl":"10.1016/j.cie.2025.111022","url":null,"abstract":"<div><div>Accurate photovoltaic (PV) power generation forecasting is critical for effective smart grid management. However, PV power is highly dependent on weather variables, resulting in significant volatility and nonlinearity in generation patterns, which makes traditional forecasting methods often struggle to guarantee both accuracy and robustness at the same time. Although Pareto-front-based ensemble learning approaches partially mitigate these issues, the excessive generation of suboptimal solutions increases decision-making challenges for grid operators. To this end, this paper customizes a novel PV ensemble learning forecasting system based on Bi-Directional Guarantee (BIDG) and Pareto Front Shrinking (PFS). Specifically, the original Hippopotamus Optimizer (HO) is enhanced into a Multi-Objective Hippopotamus Optimizer (MOHO) using a crowding-distance-based multi-objective framework. To further enhance the algorithm’s performance, low-discrepancy sequence initialization is adopted to improve the optimizer’s exploration ability, and a perturbation operator is introduced to enhance convergence performance. The BIDG mechanism based on the improved MOHO is used as the weighting strategy for ensemble learning, which imposes dual constraints on the accuracy and robustness of the forecast and penalizes suboptimal solutions to improve the quality of the Pareto front. To further improve the quality of the Pareto front to simplify decision-making, we combine the Shrinking Method to Quantify Competition (SMQC) and the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) to shrink the Pareto front. Numerical validation is conducted using seasonal local datasets from a PV plant in Australia’s Northern Territory. The experimental results show that the average absolute percentage error of the hybrid framework on the four datasets are <em>MAE_Spring</em> = 0.3518, <em>MAE_Summer</em> = 0.5689, <em>MAE_Autumn</em> = 0.8585, <em>MAE_Winter</em> = 1.2175, which are significantly lower than other models, achieving higher forecasting accuracy and demonstrating its practical applicability for real-world smart grid operations.</div></div>","PeriodicalId":55220,"journal":{"name":"Computers & Industrial Engineering","volume":"203 ","pages":"Article 111022"},"PeriodicalIF":6.7,"publicationDate":"2025-03-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143642036","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A nonparametric degradation modeling method based on generalized stochastic process with B-spline function and Kolmogorov hypothesis test considering distribution uncertainty","authors":"Zhongze He ,&nbsp;Shaoping Wang ,&nbsp;Di Liu","doi":"10.1016/j.cie.2025.111036","DOIUrl":"10.1016/j.cie.2025.111036","url":null,"abstract":"<div><div>The increasing reliability requirements of modern industrial products necessitate precise degradation modeling. Stochastic process-based methods are widely employed due to their robust uncertainty quantification capabilities but often rely on the assumption of a predefined degradation distribution, which may not hold in complex scenarios. This study presents a novel degradation modeling framework that integrates nonparametric estimation and stochastic processes into a unified approach. Unlike conventional methods combined nonparametric estimation and stochastic processes by separately modeling, the proposed method employs a nonparametric model to characterize generalized independent increment processes. Utilizing the flexibility of B-spline functions, the method effectively captures the uncertainties of degradation distributions while mitigating errors associated with improper distribution assumptions. The B-spline construction is further formulated as a numerical optimization problem supported by the Kolmogorov hypothesis test, enabling the direct determination of confidence levels for the constructed model. A Monte Carlo-based framework is employed for reliability assessment and lifetime prediction. Validation using simulations and real-world data demonstrates that the proposed method achieves superior accuracy in capturing degradation dynamics and significantly outperforms traditional methods.</div></div>","PeriodicalId":55220,"journal":{"name":"Computers & Industrial Engineering","volume":"203 ","pages":"Article 111036"},"PeriodicalIF":6.7,"publicationDate":"2025-03-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143620987","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"TAOD-Net: Automated detection and analysis of thyroid-associated orbitopathy in facial imagery","authors":"YongKyung Oh ,&nbsp;Giheon Koh ,&nbsp;Jiin Kwak ,&nbsp;Kyubo Shin ,&nbsp;Gi-Soo Kim ,&nbsp;Min Joung Lee ,&nbsp;Hokyung Choung ,&nbsp;Namju Kim ,&nbsp;Jae Hoon Moon ,&nbsp;Sungil Kim","doi":"10.1016/j.cie.2025.111024","DOIUrl":"10.1016/j.cie.2025.111024","url":null,"abstract":"<div><div>Thyroid-Associated Orbitopathy (TAO), a common autoimmune thyroid disease, significantly impacts patients’ quality of life. The conventional method for assessing TAO disease activity relies on the Clinical Activity Score (CAS), which is evaluated by skilled experts. However, the high cost of securing expert evaluators and inconsistencies in their assessments highlight the need for an expert-level, data-driven CAS assessment system. In response, we introduce TAOD-Net (Thyroid-Associated Orbitopathy Detection Network), an advanced data-driven system designed to identify five key CAS components related to inflammatory signs. Leveraging patient facial images as input, our system incorporates a novel learning strategy for multi-label classification and utilizes domain knowledge for optimized image cropping. The performance of TAOD-Net was rigorously validated using 2040 digital facial images collected from 1020 TAO patients at the Department of Ophthalmology, Seoul National University Bundang Hospital. Our results demonstrate that TAOD-Net surpasses existing models in diagnosing TAO disease activity, underscoring its potential to exceed current standards.</div></div>","PeriodicalId":55220,"journal":{"name":"Computers & Industrial Engineering","volume":"203 ","pages":"Article 111024"},"PeriodicalIF":6.7,"publicationDate":"2025-03-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143631736","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Automated argumentation-based social trust negotiation in collaborative networks","authors":"Javier Andrade-Garda ,&nbsp;Víctor Carneiro-Díaz ,&nbsp;Daniel Lage-Etchart ,&nbsp;Sonia Suárez-Garaboa","doi":"10.1016/j.cie.2025.111026","DOIUrl":"10.1016/j.cie.2025.111026","url":null,"abstract":"<div><div>Collaboration is an inherent property in Collaborative Networks (CNs), that are increasingly becoming the mainstay of manufacturing and production firms. Collaboration success largely depends on trust. Therefore, trust should be managed in an explicit manner when selecting partners to collaborate with in a CN. Traditionally, the partner selection approaches that explicitly consider trust as a selection criterion manage it as a single concept. However, trust is a multidimensional concept, and each dimension has unique influences on business relationships. The focus of the paper is precisely on the social dimension of trust, which has proven to be of special relevance in CNs. The goal is to allow partners to create alliances where this trust dimension is guaranteed, thus maximizing the chances of successful collaboration. To achieve this, an innovative automated argumentation-based negotiation framework is proposed. In this framework, partners will be modelled as agents in a Multi-Agent System, where automated argumentation techniques are widely used, and they will use different types of arguments to negotiate their social trust requirements. A prototype developed in NetLogo was used to test the proposed framework using the Spanish network on environmental DMAs (Red Española de DMAs Ambientales, REDMAAS), a Spanish thematic research network funded by the Spanish Ministry of Science and Innovation (RED2018-102594-T).</div></div>","PeriodicalId":55220,"journal":{"name":"Computers & Industrial Engineering","volume":"203 ","pages":"Article 111026"},"PeriodicalIF":6.7,"publicationDate":"2025-03-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143580835","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Manufacturer’s encroachment strategies when facing the retail platform with AI-driven pricing","authors":"Wenhui Zhou,&nbsp;Yu Ding,&nbsp;Yanhong Gan,&nbsp;Wenting Ma","doi":"10.1016/j.cie.2025.111003","DOIUrl":"10.1016/j.cie.2025.111003","url":null,"abstract":"<div><div>The rapid development of artificial intelligence technology is facilitating the adoption of AI-driven personalized pricing. Retail platforms with a higher level of technology and a closer connection with consumers have taken the lead in adopting AI-driven pricing to improve pricing flexibility. In this paper, we explore the manufacturer’s encroachment strategies in the context of the retail platform’s adoption of AI-driven pricing by constructing a game-theoretic model, as well as clarifying the impact of the magnitude of AI capability on the profits of supply chain participants. We construct three channel structures, namely, a single-channel supply chain model without manufacturer encroachment, a dual-channel supply chain model with manufacturer uniform pricing encroachment, and a dual-channel supply chain model with manufacturer AI-driven pricing encroachment. The results show that manufacturer encroachment with uniform pricing does not lead to an increase in profit. However, the manufacturer can encroach with AI-driven pricing to gain benefits when the AI capability is high, the total cost of adding a direct channel and acquiring AI technology is low, and consumers’ valuation is high. We also find that increased AI capability benefits the manufacturer in any scenario, but not for the retail platform. In the single-channel scenario, enhanced AI capability can improve the retail platform’s profit if consumers’ valuation is relatively low. However, in the dual-channel scenarios where the manufacturer encroaches, enhanced AI capability is completely detrimental to the retail platform. This also brings a management insight to retail platforms that high AI capability may not necessarily boost profitability. Finally, we also conduct a case study to substantiate our claims.</div></div>","PeriodicalId":55220,"journal":{"name":"Computers & Industrial Engineering","volume":"203 ","pages":"Article 111003"},"PeriodicalIF":6.7,"publicationDate":"2025-03-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143620986","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}