Order-picking is one of the most labor- and cost-intensive operations in warehouses, especially under the pressures of e-commerce growth and supply chain disruptions. Globally, order-picking accounts for 50–75% of total warehouse operating costs and nearly 55% of labor time, making it a dominant factor in logistics performance. Improving picking efficiency is therefore essential not only for reducing operational costs but also for enhancing resilience and sustainability in logistics. This study investigates the combined impact of storage space allocation and picker routing strategies on performance in a real-world edible oil factory warehouse with a three-block U-shaped layout. Three allocation policies (dedicated, turnover-based class storage, and family-based class storage) and three routing methods (S-shape, return, and midpoint) were tested in nine combinations over a five-week period. Results show that storage allocation has a stronger influence on picking efficiency than routing decisions. The family-based (Class 2) allocation with return routing achieved the lowest weekly picking time, reducing retrieval effort by concentrating items in low-level storage locations. Beyond efficiency gains, the findings highlight how simple, low-cost adjustments to storage policies can reduce picker travel, lower energy use, and support sustainable warehouse operations. This case study provides practical guidance for managers of small and medium-sized warehouses and offers baseline insights for the development of digital twin models and smart warehousing solutions in Industry 4.0.
This study examines magnetohydrodynamic (MHD) heat and mass transfer of a ternary hybrid nanofluid over a rotating sphere incorporating thermophoretic part
Standard operating procedures (SOPs) serve as critical control mechanisms in manufacturing systems, yet systematic approaches for quantifying procedural inefficiencies remain theoretically underdeveloped. Unlike traditional qualitative SOP analysis methods that rely on expert intuition and subjective assessment, current procedural optimization approaches lack the systematic rigor applied to physical process improvement. While lean manufacturing principles have demonstrated effectiveness in physical process optimization, their systematic application to procedural analysis represents an unexplored theoretical domain with significant potential for manufacturing systems improvement. This research addresses this gap by developing the Procedural Waste Index (PWI) framework, which establishes the first systematic theoretical integration of lean waste identification principles with procedural analysis. The framework extends the seven wastes of lean manufacturing to procedural analysis through systematic mapping to procedural elements identified via the extended Procedure Representation Language (e-PRL), creating a quantitative approach that enables the objective measurement of procedural efficiency where only subjective assessment methods previously existed. The PWI framework provides the following three key advantages over existing approaches: (1) systematic waste identification using proven lean principles rather than ad hoc improvement methods, (2) quantitative measurement capability enabling objective assessment and statistical process control, and (3) multi-perspective analytical framework through three complementary calculation methodologies (weighted aggregation, maximum constraint identification, and root mean square analysis) providing comprehensive analytical perspectives on procedural waste across discrete manufacturing contexts. The theoretical framework demonstrates practical applicability through a systematic analysis of a respirator fit testing procedure, revealing inventory waste as the primary inefficiency (70.0% waste score). This represents the first quantitative procedural waste assessment in the manufacturing literature, contributing to the foundational theory for systematic procedural optimization while establishing a methodology for future empirical validation studies.
The blood is identified as a non-Newtonian fluid. The current exploration deals with investigating blood-based radiative micropolar fluid flow of Cu-CuO hybrid nanofluids and the features of heat transfer across a movable sheet during the non-uniform heat source/sink and melting procedure. Initially, the problem is modelled in the form of partial differential equations and then altered into ordinary differential equations using similarity variables. These equations are numerically solved using the fourth-order boundary value solver bvp4c and a neural network depending on the algorithm of the Levenberg–Marquardt backpropagation. The results demonstrated that the proposed artificial neural network technique could hold non-linear data with minimum error and showed consistent performance across all phases, including training, validation, and testing. In addition, the outcomes reveal that the stretching sheet velocity is lower in contrast to the free stream velocity because of the stagnation parameter. Conversely, the temperature as well as the angular velocity decline due to the stagnation factor. Moreover, the skin-friction, the coefficient of couple-stress, and the heat transfer rate increase significantly by up to 5.71%, 7.90%, and 8.62%, respectively, with higher nanoparticle volume fractions. In contrast, the heat transfer rate decreases by about 0.0078% due to the stronger influence of the melting effect.
Modern spam is worldwide and multilingual, making advanced detection difficult for resource-constrained devices to balance with user privacy and efficiency. The lack of multilingual training data and the high computing cost of Large Language Models hinder traditional solutions. A secure, privacy-preserving framework for multilingual spam classifier development is presented and tested in this research. First, we synthetically generate a balanced, 21-language dataset by high-fidelity translating the English Enron corpus using the Aya-101 model to handle data scarcity. Next, we use Quantized Low-Rank Adaptation to effectively fine-tune the 4-bit quantized versions of leading compact LLMs Gemma 3-4B, Qwen 3-4B, and Llama 3.2-3B in a thorough comparison. QLoRA-tuned models improve significantly, with the Gemma 3-4B model performing best with 90% accuracy and 0.88 F1-score. A stunning 22-percentage-point increase above its 16-bit baseline. The fine-tuned models also reduced VRAM footprint significantly, proving their on-device feasibility. QLoRA and synthetic data pipelines provide a compelling, practical blueprint for implementing cutting-edge AI in privacy-sensitive, resource-constrained situations, according to this research.
The artificial neural network (ANN) plays a novel and innovative role in the of heat transfer by serving as a powerful predictive and data-driven modelling tool that captures complex thermal behaviors with excellent accuracy. This investigation aims to provide an optimized ANN model for oscillating flow of viscoelastic micropolar hybrid nanofluid with evaluation of heat and mass transfer. Hybrid nanofluid features characterize the suspension of aluminum alloy (AA7072-AA7075) with human blood as the base liquid. Both energy and concentration equations are modified by employing the relations of the advanced Cattaneo-Christov model. The radiated features have contributed to modifying the energy equation. Modeling of the problem leads to a nonlinear system of partial differential equations (PDEs), which are handled analytically. Based on the simulated numerical model, the ANN algorithm is implemented to predict the accuracy of the model. Understanding of machine learning algorithms has been examined through regression plots, fit functions, mean residual error, and histogram curves. Physical observations of models are revealed by evaluating the role of modeled parameters. By combining analytical simulations with ANN-driven predictive results, the analysis revealed a powerful framework for improving thermal performance, enhancing the flow stability, and enabling complex hybrid nanofluid systems. It has been observed that the developed ANN model exhibited excellent predictive capability with suitable convergence, while regression, mean square error and histogram confirmed its reliability and stable accuracy. The enhanced nanoparticles volume fraction and Brownian movement enhances the thermal impact of blood-based nanofluid. The larger micropolar fluid parameter and concentration relaxation presents a decay in the concentration diffusion.