International Journal of Industrial Engineering: Theory, Applications and Practice

Investigating Deformation and Residual Stress in Overhang Structures Produced by Laser Powder Bed Fusion: An Experimental And Thermo-Mechanical Simulation Study

Yang Liu — 2026-10-13

In laser powder bed fusion (L-PBF), overhang structures are particularly vulnerable to warping and residual stress buildup due to inadequate thermal dissipation and limited mechanical support. While contour scanning is commonly used to improve surface quality, its effects on stress distribution and deformation mechanisms in overhang geometries are not fully understood. This study systematically investigates the impact of contour scanning on overhang components with angles from 45° to 60°, combining experimental tests with coupled thermo-mechanical finite element (FE) simulations. Overhang specimens made of 316L stainless steel were produced using an SLM 125HL system. Experimental results revealed that increasing the overhang angle led to greater deformation, surface defects, and melt-through occurrences. To better understand these behaviors, an FE model incorporating Goldak’s double-ellipsoidal heat source and temperature-dependent material properties was developed to simulate the evolution of temperature and stress during contour scanning. Using a reduced-geometry approach—where the model is scaled down in all dimensions to represent three deposited layers (90 μm) over a 1 mm scan length, with volume hatching excluded—the simulation successfully replicates the qualitative deformation patterns and residual stress accumulation observed in the experiments. The simulation further reveals that localized thermal gradients around contour paths induce asymmetric warping at the free ends of overhangs and generate stress concentrations near constrained regions. These insights emphasize the pivotal role of contour scanning in the formation of deformation and residual stresses and provide valuable guidance for optimizing processing parameters to successfully fabricate unsupported overhang features in L-PBF.

Mixed-Integer Linear Programming Model for Integrated Liner Shipping Route and Schedule Optimization with Weekly Port Time Windows

Xi Jiang — 2026-10-13

Designing an efficient and cost-effective port network requires accounting for practical constraints, such as port time operation windows. Ignoring these constraints can lead to impractical schedules. This paper presents a method for optimizing liner shipping routes and schedules while accommodating multiple weekly port time windows. A mathematical model is developed to minimize overall costs by optimizing the sequence of port calls, schedules, and sailing speeds. The model is linearized as a mixed-integer program and validated through numerical experiments using commercial optimization solvers. A case study based on SITC Container Lines validates the proposed model and confirms its cost-saving effectiveness, showing total cost reductions of 13.28%, 32.74%, and 16.26% across three port time window scenarios compared with the original port call sequence. Sensitivity analyses further examine the effects of bunker prices, port time windows, and shipping demand volume.

HIPA: A Human Information Processing Architecture for LLM Responses in Ergonomic Design

Youngjun Park — 2026-10-13

Large language models (LLMs) are increasingly used for UI/UX evaluation and design decision support; however, their responses are often generated as implicit and unstructured outputs, making cognitive processes and judgment rationale difficult to trace. Moreover, the quality of LLM-based judgments may vary depending on users’ prompting proficiency, limiting the consistency of AI-assisted evaluation. This study proposes HIPA (Human Information Processing Architecture), a cognitive response architecture that structures LLM outputs based on the human information processing (HIP) model and investigates the manifestation of human cognitive stages in conventional LLM responses. HIPA maps four HIP stages—perception, working memory, decision making, and response execution—into a two-stage response structure consisting of information provision and judgment support through knowledge-grounded prompting with GPT-4o. An exploratory evaluation was conducted with seven participants with different prompting proficiency levels who assessed the same mobile login interface using a general LLM. Responses were coded according to HIP-stage manifestation using a three-point scale. The results showed that perception and response execution were consistently observed, whereas working memory and decision-making stages varied depending on prompt systematicity. These findings indicate that general LLM responses do not inherently provide cognitively structured judgment support. HIPA provides a human-centered framework for developing AI-assisted UI/UX evaluation systems based on human cognitive principles.

Understanding Human-AI Trust in Military AI Applications: A Multistage Human Factors Framework

Kangwon Kim — 2026-10-13

Artificial intelligence (AI) is increasingly used in military systems to support intelligence, surveillance, reconnaissance, autonomous operations, logistics, command and control, and decision support. Effective use, however, depends not only on technical performance but also on how operators perceive, interpret, evaluate, and rely on AI outputs under uncertainty, time pressure, and adversarial interference. This study presents a conceptual integrative review informed by a structured supplementary literature search and synthesis. It synthesizes human factors, human–automation trust, human–AI teaming, military AI, and industrial engineering literature to develop an iterative multistage framework of human–AI trust in military applications. Trust is conceptualized as dynamic situational trust in a specific military AI system during task interaction, while dispositional, initial, and learned trust are treated as antecedent conditions. The framework distinguishes four analytically separable but functionally interdependent stages—perception, interpretation, evaluation, and trust judgment and reliance decision—and identifies four military-relevant moderators: system transparency, explainability, and behavioral predictability; data quality and information uncertainty; adversarial vulnerability; and operational and organizational context. The framework clarifies conceptual pathways to calibrated reliance, overtrust, and undertrust, and derives implications for task analysis, interface evaluation, human-centered design, operator training, and empirical validation.

Impact of Visual Aids on Posture And Musculoskeletal Risk Among Dental Practitioners: A Human Factors Approach

Rahmah Abdulsalam Attallah — 2026-10-13

Work-related musculoskeletal disorders are highly prevalent among dental practitioners due to sustained static postures and limited visual accessibility. This study examined the effects of three visual aid conditions: naked eye, magnification loupes, and dental operating microscope on operator posture and musculoskeletal risk using a human-factors approach. Ninety practitioners (n = 30 per group) performed standardized clinical tasks, with posture captured using multi-angle high-resolution photography. Ergonomic risk was assessed using the Rapid Upper Limb Assessment, Rapid Entire Body Assessment, and Branson’s Posture Assessment Instrument, and analyzed using one-way analysis of variance with post hoc comparisons. Significant differences were observed across all measures: RULA (F(2,87)=49.14, p<0.001), REBA (F(2,87)=51.97, p<0.001), and PAI (F(2,87)=37.20, p<0.001). The microscope condition showed the lowest ergonomic risk scores (3.10, 2.80, and 3.23, respectively) compared with the naked-eye and loupes conditions. These findings suggest that visual magnification may contribute to improved ergonomic posture, particularly in the neck and trunk regions, and may help reduce musculoskeletal risk during dental procedures.

A Lean Approach to Improving the Laboratory Verification Process in the Electronics Industry

Ping-Shun Chen — 2026-10-13

This study investigates the use of lean thinking to improve process efficiency in the mechanical stress test laboratory of Company I, a semiconductor verification firm in Taiwan. It addresses a practical problem in which, despite relatively stable project volumes, project working hours increased due to greater case complexity, inefficient project handovers, and operational waste in laboratory workflows. The study introduces a modified 6Ws framework, replacing the conventional “how” dimension with “whom” to strengthen problem definition in a service-intensive laboratory context. This framework was combined with a suppliers, inputs, processes, outputs, and customers (SIPOC) analysis and value stream mapping (VSM) to identify non-value-added activities, redesign project intake and handover procedures, improve scheduling visibility, and optimize workspace and tool arrangements. The results show that the controllable labor component was reduced from 160 minutes to 58 minutes, representing a 63.75% reduction in labor time. The 102-minute improvement is a substantial reduction in human effort across the 50 hours of the total end-to-end cycle, which is dominated by machine operation and testing. Overall, the findings demonstrate that lean thinking can significantly reduce avoidable labor waste in laboratory operations and provide a practical improvement model for similar technical service environments.

A Hybrid Genetic Algorithm–Tabu Search Method for The Storage Location Assignment Problem in Robotic Mobile Fulfillment Systems

Yung-Chia Chang — 2026-10-13

With the advent of artificial intelligence, the robotic mobile fulfillment system (RMFS) has emerged as a novel automated warehousing system. Using a goods-to-person approach, RMFS employs robots to transport pods to workstations, where human pickers retrieve the required items, thereby significantly reducing walking distance. In addition, both material assignment to pods and pod storage location assignment play critical roles in improving picking efficiency. This study addresses these two decision problems through a two-stage solution framework under a simplified single-AGV, single-workstation RMFS. In the first stage, Apriori-based association analysis and DBSCAN clustering are employed to identify highly associated materials and assign them to pods. In the second stage, a hybrid genetic algorithm (GA) and tabu search (TS) algorithm is developed to optimize pod storage locations by minimizing the total round-trip picking distance required to fulfill orders. To verify the feasibility and effectiveness of the proposed approach, the developed algorithms were applied to one week of real production-order data from a Taiwanese computer equipment manufacturing company. Experimental results show that, compared with the company's current method, the proposed approach reduces the average total pod movements by approximately 47.12% and the average total picking distance by approximately 29.60% under the studied operational setting.

Routing Optimization for The Traveling Salesman Problem with Drones (TSP-D) in Medical Supply Distribution

Yutong Ha — 2026-10-13

The integration of drones into last-mile logistics has emerged as a transformative solution to address the inefficiencies of traditional truck-based delivery in smart cities. This study investigates the Traveling Salesman Problem with Drones (TSP-D) specifically within the context of medical supply distribution in Qingdao, China. A hierarchical two-stage optimization methodology is proposed: first, the service area is partitioned into optimal subregions using the k-means clustering algorithm, with the number of clusters determined by the silhouette coefficient. Second, a mixed-integer programming model is formulated and solved using Gurobi to optimize synchronized routing within each cluster. Based on a dataset of 169 customer locations, computational experiments demonstrate that the framework identifies seven efficient clusters and significantly enhances delivery performance. The findings indicate that the proposed framework has the potential to reduce response time and improve operational efficiency in medical supply distribution.

Integrated Modeling, Simulation, and Experimental Optimization of Vibratory Coffee Sorting Performance Using Response Surface Methodology

Engelbert Harsandi Erik Suryadarma — 2026-10-13

This study presents an integrated theoretical, simulation, and experimental approach to optimize the performance of a vibratory coffee sorting machine using Response Surface Methodology (RSM). A physics-based model of granular flow and segregation was developed to quantify the influence of inclination angle (α) and vibration frequency (f), and numerical simulations using MATLAB and DEM confirmed nonlinear interactions between both parameters, predicting optimal performance at α = 28.5° and f = 47.5 Hz with a throughput of 3541.3 g/min and sorting accuracy of 98.14%. Empirical validation through a Central Composite Design (CCD) demonstrated significant effects of α and f on system responses (p < 0.001) and produced highly accurate regression models (R² = 98.32% for throughput and R² = 92.52% for accuracy). Multi-response desirability optimization yielded an empirical optimum at α = 29.8° and f = 50 Hz, achieving a throughput of 3667.44 g/min and an accuracy of 98.32%, closely aligning with simulation predictions. The results confirm the robustness of the integrated modeling framework, highlight the inherent trade-off between throughput and segregation efficiency, and provide a practical guideline for improving productivity and quality in vibratory coffee sorting operations.

A Cluster-Based Technology Roadmapping Approach for Strategic Management of Artificial Intelligence in Healthcare

Sakire Nesli Demircioglu — 2026-10-13

Technology roadmapping is a key strategic concern for companies. The aim of this study is to propose a cluster-based technology roadmapping approach using patent documents. The proposed approach is applied to artificial intelligence (AI) technology. The proposed approach provides evidence-based decision-making in healthcare management by identifying and prioritizing artificial intelligence (AI) technology domains with high long-term success potential. Unlike existing studies in the literature, the proposed approach integrates weighted evaluation criteria with market trends and product feature characteristics, enabling a more comprehensive assessment of technological relevance from a managerial perspective. The applicability of the proposed approach is demonstrated through an empirical case study on AI applications in the healthcare domain. The findings suggest that medical image processing, diagnosis and treatment support systems, electronic communication technologies, interactive robotic systems, and acoustic and haptic feedback technologies are expected to play a significant role in the long-term development of healthcare services.

Human-Robot Collaboration in Manufacturing: A Systematic Review of Current State, Evolutionary Trajectory, and Integration Gaps

Kung-Jeng Wang — 2026-10-13

Driven by the imperative to enhance manufacturing flexibility and resilience, this study undertakes a comprehensive assessment of human-robot collaboration (HRC) technology. Its aims are threefold: to evaluate the current technological landscape, delineate its evolutionary trajectory, and pinpoint critical gaps in existing HRC implementations. The ultimate goal is to synthesize these findings into an integrated technology-effect-application reference map that researchers and practitioners can use to identify priority areas for HRC development, rather than a prescriptive integration framework. Our methodology integrates citation and cluster analysis within a systematic literature review of 121 peer-reviewed articles. Main path analysis was selected over a purely qualitative synthesis because it traces citation-based evolution using a reproducible, quantitative weighting criterion, while K-modes clustering was selected over alternatives such as K-means because it natively handles the categorical (rather than continuous) nature of the technology-effect-application coding. Citation analysis revealed HRC as a multifaceted, multi-field technology still awaiting the formation of an interconnected knowledge system. Subsequently, K-modes-based cluster analysis identified three distinct subgroups across technology, effect, and application dimensions. The analysis yielded a structured taxonomy encompassing 14 unique HRC technologies, 16 associated technological effects, and five industrial applications. Citation network analysis further exposed 11 discrete research clusters, signaling limited cross-disciplinary integration and the nascent state of HRC research interconnectivity. A two-dimensional matrix analysis pinpointed five predominant technologies propelling HRC advancement: data analytics systems, optimization and machine learning algorithms, machine vision systems, sensor networks, and human-machine interfaces. Critically, this analysis simultaneously uncovered significant technological deficiencies, notably in safety protocols and cognitive interaction frameworks, indicating areas ripe for targeted research. This research significantly contributes to the HRC domain by elucidating the contemporary state of HRC technology, charting its developmental path, and comprehensively highlighting persistent research deficiencies.

A Hybrid Deep Recommendation System for Issuing Digital Coupons in Real-Time

Wongyeom Seo — 2026-10-13

In the field of marketing, offering discount coupons to encourage purchases is one of the most frequently used strategies. In particular, with the rise of online shopping malls surpassing offline ones, digital coupons have become increasingly prevalent. The key challenge in issuing coupons is identifying customers with a high risk of churn and providing them with coupons of the right category at an optimal discount rate. Manually addressing this task is impractical because online shopping platforms serve millions of users and offer a wide variety of coupon types. Accordingly, this study proposes a time series deep learning framework that identifies customers at high risk of churn in real time and provides them with personalized coupons for relevant product categories at appropriate discount rates. To address the data class imbalance problem, additional dependent variables were introduced, and static customer meta-information was combined with real-time page click features to enhance accuracy. In addition, we tested the improvement in financial profits of an actual shopping mall. After various ablation model experiments, the results show the superiority of our proposed framework, which has brought great profits to the platform. The proposed framework achieved an AUC of 0.747 for purchase intention prediction, improved coupon conversion rates by up to 32.65%, and increased the relative profit index from 100.0 to 103.15 in real-world deployment experiments.

Prediction of Material Removal Rate for Chemical Mechanical Planarization using Bayesian Additive Regression Trees

Han Bao — 2026-10-13

Chemical Mechanical Planarization (CMP) is a critical surface processing technique in semiconductor manufacturing, where the Material Removal Rate (MRR) plays a vital role in determining process quality. Traditional physics-driven MRR models based on Preston's Equation have difficulty capturing the complex nonlinear relationships and high-order interactions present in CMP data. This paper applies Bayesian Additive Regression Trees (BART) to CMP MRR prediction. BART uses a Bayesian sum-of-trees framework to model nonlinear effects and interactions while providing posterior uncertainty estimates and variable-importance information. Quantitative comparisons on the validation set show that BART achieves the best predictive performance, outperforming the strongest baseline model by 11% in terms of RMSE. Model reliability and interpretation are examined through residual diagnostics, MCMC convergence checks, hyperparameter-sensitivity analysis, omnibus covariate testing, and variable-importance analysis. These results indicate that BART is effective for CMP MRR prediction and can help identify process variables that warrant monitoring and diagnostic attention.

An Adaptive Optimization Strategy Applied to the Artificial Bee Colony Algorithm

Oussama Ait Sahed — 2026-10-13

Balancing exploration and exploitation remains a central challenge in metaheuristic optimization, particularly for engineering applications requiring both global search efficiency and reliable convergence. This paper introduces a simple and adaptive strategy that dynamically adjusts the behavior of an optimization algorithm based on real‑time performance feedback. The method identifies key search operators and assigns each an explorative and exploitative variant, grouped into two optimization modes. An online mechanism switches between these modes according to the successful mutation rate, enabling the optimizer to respond to the evolving landscape of the search process. The strategy is integrated into the artificial bee colony algorithm and evaluated on 18 numerical benchmarks, a CEC2021 test suite, and a nonlinear model predictive control problem. Results demonstrate improved robustness, competitive accuracy, and faster convergence by reducing the number of evaluations by more than 4% against the established GABC, highlighting the method’s suitability for complex industrial optimization

Optimal Selling and Leasing Strategy for A Dominant Durable Goods Manufacturer Under Sequential Market Entry and Rental Platform Competition

Wei Xiao — 2026-10-13

Recent advances in digital and information technologies have significantly enhanced the attractiveness of product leasing through online rental platforms, especially in durable product markets. Prominent manufacturers such as BMW, IBM, GM, and Epson have embraced the leasing model as a means to promote their products. However, limited studies have addressed the impacts of a competitor’s encroachment on the strategic decisions of the dominant manufacturer. In this paper, we develop a two-period dual-channel analytical framework where the dominant manufacturer exists in period 1, and an entrant encroaches on the market share in period 2. Our analysis shows that, no matter how the market size changes, the dominant manufacturer can always benefit from the leasing channel, and it is never in its best interest to adopt a pure leasing strategy in period 1. If the market size shrinks in period 2, the dominant manufacturer prefers selling products directly to consumers in period 1, resulting in a decline in leasing ratios. Moreover, our analysis indicates that the entrant’s encroachment may increase the dominant manufacturer’s profit under specific conditions.

Nexsupply: A Multi-Objective Sequential Decision Framework for Spare-Part Inventory Optimization Under Demand and Lifecycle Uncertainty

Rong-Guei Tsai — 2026-10-13

Inventory management of maintenance spare parts is essential for the reliability and cost efficiency of urban water supply systems where heterogeneous components experience intermittent failure and demand. Conventional replenishment policies minimize procurement cost alone and overlook the operational overhead of frequent ordering. In this paper, we propose NexSupply, a finite-horizon sequential decision framework for spare-part inventory optimization under demand and equipment-lifecycle uncertainty. NexSupply formulates replenishment as an exact dynamic program over a per-material (period, inventory) state space and internalizes the ordering-frequency objective directly into the state-transition cost by weighted-sum scalarization. The frequency-penalty weight lambda trades off operational cost against the number of replenishment actions. Lambda equal to zero recovers the cost-only policy exactly. Experiments on synthetic instances and real-world cases against cost-priority, management-priority, EOQ, genetic-algorithm, and single-objective dynamic-programming baselines show that NexSupply achieves near cost-optimal total cost (which is indistinguishable from the cost-only optimum) while reducing ordering days by roughly 90% compared to a cost-only policy (which is less than 1% more cost). The dynamic program scales linearly in the planning horizon and number of materials, solving realistic cases in milliseconds. NexSupply provides a practical and tunable decision-support tool that balances economic efficiency and managerial overhead for uncertainty-stacked decision-support tools.

Selling-Mode Choice in Live-Streaming E-Commerce Under Platform Traffic Pricing and Heterogeneous Consumer Preferences

Yingying Li — 2026-10-13

The rapid growth of live-streaming e-commerce has raised important questions regarding how suppliers should select selling modes when collaborating with live-streaming enterprises (LSEs) in the presence of heterogeneous consumer preferences and platform traffic pricing. We develop a game-theoretic model to examine the strategic interactions among a platform, a supplier, an LSE, and two types of consumers: shopping-preferred (SP) consumers and content-preferred (CP) consumers. The model examines how LSE selling ability, consumer preferences, and platform traffic pricing jointly affect selling-mode choices, traffic decisions, profitability, and the strategic alignment of preferences between the supplier and the LSE. Three main findings emerge. First, selling-mode preferences depend primarily on commission incentives and consumer preferences. Agency selling can align the supplier’s and the LSE’s preferences when revenue sharing provides sufficient traffic acquisition incentives without imposing excessive commission costs; otherwise, reselling may be mutually preferred. Second, both the supplier and the LSE prefer targeting CP consumers under either selling mode because livestream content creates additional consumer value and increases the sales return generated by acquired traffic. Third, the platform’s equilibrium traffic price is determined by its marginal traffic-service cost, whereas the equilibrium amount of traffic acquired varies with the commission rate, LSE selling ability, and consumer preferences. An extension incorporating consumer viewing costs shifts the relevant selling-mode preference boundaries but does not overturn the main qualitative findings. These results provide managerial guidance for selling-mode selection, consumer targeting, and platform traffic pricing in live-streaming e-commerce.

The Digital–Circular Economy Nexus in Supply Chains: Trends, Themes, and Insights from a Scientometric Perspective

Muhammad Shoaib — 2026-10-13

The growing urgency for sustainability in global supply chains, spurred by increasing demand, environmental degradation, and resource scarcity, has intensified academic and industrial focus on adopting circular economy (CE) principles. Recently, digital technologies have become key enablers of this shift, facilitating improved supply chain resilience, efficiency, and environmental outcomes. This study addresses a critical gap in the literature by systematically exploring the convergence of digitalization, supply chain management, and the circular economy. A scientometric analysis of 628 peer-reviewed articles published between 2010 and 2025 offers a comprehensive mapping of the intellectual structure and evolution of knowledge in this field. The findings reveal a marked acceleration in publication volume in 2025 compared with previous years, particularly in journals focusing on environmental sustainability and technological innovation. Citation and network analyzes indicate that China dominates research output, whereas the United Kingdom and India exhibit higher citation impact, reflecting their intellectual prominence in the field. Content analysis reveals four core thematic clusters: the technological influence of digital tools such as blockchain, the Internet of Things, artificial intelligence, digital twins, and machine learning; strategic approaches to strengthening supply chain resilience; emerging opportunities and innovation potential; and real-world industrial deployment across various sectors. These themes collectively underscore a growing scholarly consensus on the integrative role of digital technologies in advancing circular economy goals within supply chain systems. By tracing the evolution of research and highlighting key authors, journals, and thematic trends, this study offers critical insights for both academia and industry. This study is grounded in Dynamic Capabilities Theory (DCT), which explains how organizations purposefully identify opportunities, seize them, and reconfigure resources to sustain competitive advantage. It provides a strategic foundation for practitioners to evaluate digital maturity and align technological adoption with circular-economy objectives. Ultimately, this study highlights the transformative role of digitalization in integrating circular principles into global supply chains, offering deeper insights into sustainable value creation in the digital era.