Engineering and Applied Science Research

CFD-Guided design and experimental validation of a horizontal ribbon-blade mixer for commercial Leucaena leucocephala-based ruminant feed production

Mon, 05 Oct 2026 00:00:00 +0700

Homogeneous mixing of compound animal feed is a prerequisite for nutritional consistency, growth performance, and economic feasibility in commercial ruminant production. Acacia (Leucaena leucocephala) wood chips have emerged as a low-cost, locally available protein-rich roughage in tropical Southeast Asia, but their irregular particle geometry and density mismatch with conventional concentrates pose a persistent processing challenge. This study presents a coupled computational and experimental framework for the design and validation of a 1.5 m³ horizontal ribbon-blade mixer tailored for L. leucocephala-based feed production. Three blade geometries—screw, paddle, and double helical ribbon—were comparatively evaluated using a transient three-dimensional CFD model (Volume-of-Fluid multiphase coupled with a moving reference frame and the k–ε turbulence closure) at shaft speeds of 10, 20, and 50 rpm over a 20 min mixing cycle. The ribbon geometry consistently outperformed the screw and paddle blades, reducing the predicted Coefficient of Variation (CV) from 42.1 % to 11.7 % within 6 min. A full-scale prototype with a centerline ribbon diameter of 1.14 m, pitch of 0.51 m, and a 10 kW motor was subsequently fabricated. Experimental sampling at ten radial–axial locations and one discharge point yielded a minimum CV of 7.01 % and a Lacey mixing index of 0.987 at 10 rpm after 5 min, in close agreement with the CFD prediction (CV = 7.1 %; relative error = 1.3 %). The validated prototype achieved a throughput of 4.5 t h⁻¹ at a specific energy consumption of 0.42 kWh t⁻¹ and a simple payback period of 2.63 years. The combined numerical–experimental–economic evidence demonstrates that CFD-guided geometric selection can deliver a commercially deployable, energy-efficient mixer for heterogeneous lignocellulosic–protein feed mixtures.

Structural and morphological characterization of tea-waste-derived carbon/TiO2 (AT–TiO2) composites prepared by solvent-assisted solid-state dispersion

Thu, 22 Oct 2026 00:00:00 +0700

Tea waste is an abundant agro-industrial residue but its valorization into functional inorganic composites remains largely unexplored. In this study, the waste was torrefied at 230 °C and combined with commercial TiO₂ by a solvent-assisted solid-state dispersion route followed by calcination at 500 °C, giving a Tea-Waste-Derived carbon/ TiO₂ material denoted AT– TiO₂. A total of 3 AT: TiO₂ mass ratios (9:1, 8:2 and 7:3) were prepared with pure AT and pure TiO₂ controls processed identically, and characterized by XRD and SEM–EDS. XRD results showed that single-phase rutile retained its structure in every composite, with no peak shift and secondary phase. The carbon content was obtained in EDS pattern, due to the effect of using ethanol as a composite solvent. Scherrer analysis of the (110) reflection gave apparent crystallite sizes of 61, 34, 61, and 49 nm for pure TiO₂ as well as the 9:1, 8:2 and 7:3 composites. SEM image analysis of 60 objects per sample, fitted to a log-normal distribution, gave median agglomerate sizes of 11.99 ± 0.51, 2.96 ± 0.32, 9.91 ± 0.74, 5.97 ± 0.66 and 11.00 ± 0.81 µm respectively. The 8:2 composite showed both the smallest median and the narrowest distribution (geometric standard deviation 1.56), indicating the most uniform dispersion. This study can serve as the basis for the development of biomass–inorganic composite materials with the potential to be applied in the fields of environmentally friendly adsorption and photocatalysis.

Field validation of a proximity-gated AI framework for mounting-based estrus detection in dairy cattle

Watthana Pongsena, Prakaidoy Ditsayabut, Phakpoom Chinprutthiwong — Mon, 05 Oct 2026 00:00:00 +0700

Accurate estrus detection remains challenging because activity-based monitoring systems may misinterpret non-reproductive movements as estrus-related activity. This study hypothesized that mounting-based estrus detection can be improved by conditioning accelerometer-based motion classification on inter-animal proximity, where mounting is physically feasible. We developed and field-tested a proximity-gated AI framework integrating Bluetooth Low Energy (BLE) proximity sensing with tri-axial accelerometer-based motion analysis. Unlike conventional activity-only ML/DL approaches, the proposed framework treats estrus detection as an interaction-dependent behavioral event by first validating close-contact feasibility and then applying motion classification and event-level aggregation. Controlled evaluation using 120 manually annotated mounting events achieved an event-level F1-score of 93.2% with the LSTM model. A retrospective ablation analysis showed that BLE proximity gating reduced false-positive event candidates from 48 to 6 compared with ungated motion-only inference, corresponding to an 87.5% reduction. Precision increased from 70.0% to 94.8%, while recall decreased slightly from 93.3% to 91.7%. The system was subsequently deployed across three smallholder dairy farms using 71 collar-mounted devices. During field deployment, 31 confirmed estrus events were recorded, of which 23 were detected, corresponding to a field-level recall of 74.2%; no false-positive alerts were observed within this limited dataset. Among 13 system-detected cows that underwent artificial insemination, 11 did not return to estrus during the first expected post-insemination cycle follow-up, giving a first-cycle non-return proportion of 84.6% (11/13), used as a field proxy rather than confirmed pregnancy. These findings support the hypothesis that proximity-gated behavioral validation improves the biological specificity of mounting-based estrus detection, although broader multi-site and longitudinal validation remains necessary.

Time-Series modeling of supply chain disruption drivers in ASEAN agri-food logistics networks

Salang Musikasuwan, Nguyen Tuong Chau, Tran Văn Binh — Mon, 19 Oct 2026 00:00:00 +0700

Disruptions in agri-food supply chains pose logistical and operational challenges for ASEAN member economies, whose agricultural sectors are simultaneously exposed to fluctuations in freight costs and climate-related production shocks. Prior time-series applications in shipping markets, agricultural commodity co-movement, and supply chain performance have generally examined these channels separately, and the joint transmission of these channels into ASEAN agri-food commodity logistics outcomes remains largely unexamined. This study applies a Vector Autoregression (VAR) time-series modeling framework to identify and quantify the principal drivers of disruption in ASEAN agri-food logistics networks, using 4,108 daily observations spanning January 2015 to March 2026. Seven variables are jointly modeled: Baltic Dry Index daily change (D_BDI), Brent crude oil log-return (R_Brent), Geopolitical Risk Index daily change (D_GPR), Oceanic Niño Index anomaly change (D_ONI), ASEAN aggregate agricultural production return (R_ASEAN_Production), and daily price changes of Robusta and Arabica coffee. Stationarity is verified by the Augmented Dickey-Fuller (ADF) tests. The lag order p = 2 is selected based on BIC and HQIC. GPR shocks exert statistically significant lagged effects on Arabica price (p = 0.007) and Robusta price (p = 0.021); ONI anomaly changes significantly propagate into Robusta prices (p = 0.021). A cross-domain pathway from ONI to GPR (p = 0.010) is consistent with climate disruption signals coinciding with subsequent geopolitical risk movements. BDI first-order persistence coefficient is 0.596 (p < 0.001). Forecast Error Variance Decomposition (FEVD) shows Robusta-to-Arabica cross-variance at 44.03% at horizon 10. Durbin-Watson statistics (1.958–2.061) confirm residual white noise across all equations. These results provide quantitative inputs for engineering early-warning systems for disruption and logistics resilience planning in ASEAN agri-food supply networks.

Accuracy assessment of BIM 5D-based quantity take-off for concrete and reinforcement works: A case study of a three-story building project

Fendi Hary Yanto, Widi Hartono , Taufiq Lilo Adi Sucipto — Tue, 20 Oct 2026 00:00:00 +0700

Accurate quantity take-off is essential for construction cost estimation, procurement, and project control, particularly for reinforced-concrete structural works. This study evaluates the accuracy of BIM 5D-based quantity take-off generated by Tekla Structures and Cubicost for concrete and reinforcement works in a three-story building project in Surakarta, Indonesia. A quantitative comparative case-study design was applied using the same structural drawings, reinforcement details, manual calculation recapitulations, and consultant bill of quantities (BOQ). Manual calculation was treated as a controlled analytical reference rather than an error-free ground truth, while the consultant BOQ was used as a secondary practical benchmark. The analysis used absolute deviation, percentage deviation, mean absolute percentage error (MAPE), root mean square error (RMSE), mean bias, Bland-Altman limits of agreement, and R² as supporting exploratory indicators. Based on the corrected element-level quantities, Tekla Structures and Cubicost produced concrete deviations of 1.803 m³ (0.30%) and 3.341 m³ (0.55%), respectively. For reinforcement, Tekla Structures and Cubicost produced deviations of 9,879.109 kg (9.07%) and 1,379.923 kg (1.27%), respectively. Deviations relative to the consultant BOQ ranged from 11.83% to 12.36% in the project-documentation comparison. The findings indicate that concrete quantities were relatively stable because they were primarily geometry-based, whereas reinforcement quantities were more sensitive to modelling assumptions, detailing rules, and extraction settings. The contribution of this study is a case-specific BIM 5D validation framework that separates analytical accuracy from BOQ agreement and highlights the importance of work-item sensitivity in structural quantity estimation.

Impact of vehicle type proportion and speed bump on traffic characteristics and emission

Wed, 26 Aug 2026 00:00:00 +0700

The installation of speed bumps on local roads in Indonesia is one solution to reduce vehicle speeds. Motorcycles and passenger car are the most common types of vehicles on local roads. The use of speed bumps can have environmental impacts due to traffic activity. The purpose of the research is to determine differences in the influence of the proportion of vehicle types on speed, fuel consumption, and emissions in conditions on and before speed bumps. Methodology of research is collecting traffic flow and speed census data by vehicle type for 6 days at the road segment point before the speed bump and on the speed bump location. The guidelines used in the research analysis are based on the Regulation of the Minister of Transportation of the Republic of Indonesia No. PM 14 of 2021 concerning Road User Devices and Safety, and the 2010 Environmental Guidelines, Indonesia. Principal results is when a vehicle travels over a speed bump, there is a significant change in speed, especially for motorcycles. Compared to motorcycles, the increased proportion of passenger car results in a more sensitive increase in fuel consumption and emissions. Contributions to the field is provide vehicle speed management especially on local road with many residential areas to improve air quality from the traffic perspective. Research limitation is limited to only two types of vehicles, namely motorcycles and passenger cars. Practical implication is can reduce climate damage caused by traffic activities, road user safety and social implication is improved safety and air quality and discipline in traffic. Vehicle speed restriction management on local roads with speed bumps to reduce emissions and fuel consumption usage is a maximum of 5.981 km/hour.

HSEL: An adaptive rotational hybrid stacking ensemble with exponential accuracy weighting for real-time forest fire risk prediction

Yuda Irawan, Sarjon Defit, Rini Sovia — Thu, 17 Sep 2026 00:00:00 +0700

This study proposes the HSEL model to improve the accuracy of fire risk predictions based on meteorological data and real-time IoT integration. The main issues with individual models are their susceptibility to bias and overfitting, while conventional stacking methods have not been able to effectively optimize the contribution of each base model. HSEL addresses these issues through the application of exponential accuracy weighting with parameter α = 5 and hyperparameter optimization using Optuna. The results show that accuracy improved from 87.80%–88.26% in the base stacking to 92.55%–94.20% after applying weighting, and reached 96.53% after optimization. Additionally, the model demonstrates stable performance, is capable of handling data imbalance, and possesses good generalization ability. This model also successfully improves classification balance across classes and significantly reduces prediction errors in the minority class. SHAP-based Explainable AI (XAI) analysis shows that the model is capable of providing transparent interpretations of the main factors influencing predictions.

Optimization of continuous biodiesel production from refined palm oil using a rotating inert packed-bed reactor

Thu, 15 Oct 2026 00:00:00 +0700

This study investigates the application of a rotating inert packed-bed reactor (RIPB) for continuous production of biodiesel from refined palm oil (RPO) using sodium hydroxide (NaOH) as a homogeneous catalyst. The study examines the impact of catalyst loading, methanol-to-oil molar ratio, rotational speed and feed flow rate on biodiesel (or fatty acid methyl ester, FAME) yield. A central composite design (CCD) approach with response surface methodology (RSM) was used to identify the optimal conditions for achieving maximum biodiesel yield. A FAME yield as high as 97.9% was achieved at a NaOH loading of 1.258 wt.% of oil, a methanol-to-oil molar ratio of 5.5:1, a rotational speed of 1,280 rpm, a feed flow rate of 210 mL/min, and a reaction temperature of 60 °C. The performance of the RIPB was further compared with other intensified reactors including the rotating tube reactor, household hand blender, and spinning disc reactor. The RIPB demonstrated the short residence time (<5 s) and the high yield efficiency (20.9 × 10⁻⁴ g/J). The measured properties of the purified biodiesel were within their corresponding limits specified in EN 14214 and ASTM D6751. These results demonstrate the potential of the RIPB for intensified continuous biodiesel production.

Tomato ripeness classification with weight-based and texture-adaptive pooling in CNNs

Wed, 07 Oct 2026 00:00:00 +0700

This research introduces a weight-based mixed pooling and adaptive fuzzy pooling mechanism based on feature map texture analysis for attention mechanism Squeeze-and-Excitation Networks (SE) and Convolutional Neural Networks (CNNs) in tomato ripeness classification. Traditional CNNs and SE networks rely on deterministic pooling strategies that can be suboptimal for capturing diverse visual features such as texture and color gradients in fruit. A novel fuzzy pooling system is proposed that dynamically adjusts the fusion ratio of min and max pooling based on real-time texture analysis of feature maps via a Laplacian filter. This mechanism was integrated into Darknet53, ResNet34, and ResNet101 architectures, both directly and within the SE attention block. The method was evaluated on a five-class tomato ripeness dataset using 10-fold cross-validation against traditional and fixed-weight pooling methods. The fuzzy and mixed pooling approach demonstrated superior performance on certain models which include ESEnet: Darknet53 + SE (Fuzzy pool) achieved an average F1 score across folds of 0.8429, outperforming Darknet53 + baseline SE (Average pool) at 0.8350; Extended Net: ResNet34 + Fuzzy pool achieved an average F1 score across folds of 0.8305, outperforming baseline ResNet34 at 0.8130 and ResNet101 + Fuzzy pool achieved an average F1 score across folds of 0.8284, outperforming baseline ResNet101 at 0.8267. This improvement was realized with minimal increase in computational overhead. Specifically, measurements of model parameters (M) and FLOPs (GMACs) showed only a slight increase across all architectures after integrating SEnet into the system, and zero increase after integrating the Laplacian filter and fuzzy logic. The study concludes that logic-based adaptive pooling is a robust strategy that improves feature representation in CNNs, achieving a balance between classification accuracy and computational cost. These findings offer practical value for developing more reliable automated harvesting systems and provide a generalizable enhancement for a wide range of computer vision classification tasks.

Drying holy basil using combined microwave and hot-air technology: An analysis of drying kinetics, energy efficiency, and color retention

Thu, 15 Oct 2026 00:00:00 +0700

Holy basil (Ocimum tenuiflorum) is a perishable herb with a high moisture content. Traditional hot-air drying methods result in longer drying times, higher energy use, and reduced quality. This study investigates combined hot-air and microwave drying to improve drying efficiency, reduce energy consumption, and maintain product color. A drying method was developed to evaluate the effects of hot air temperature (60, 70, and 80 °C) and microwave power level (300, 400, and 500 W) on drying kinetics, moisture diffusivity, color retention, and energy consumption. A mathematical transport model was developed in COMSOL Multiphysics to simulate the drying process. The model showed good agreement with experimental data. The results showed that microwave-hot air drying substantially accelerated moisture removal. Higher microwave power and hot-air temperature reduced drying time and specific energy consumption while increasing the effective moisture diffusivity. Combined hot‑air and microwave drying resulted in a lower total color change. The highest effective moisture diffusion coefficient (2.79 × 10⁻¹¹ m²/s) and the lowest total specific energy consumption (529.91 MJ/kg _{water evaporated}) were obtained at a hot-air temperature of 80 °C and a microwave power of 300 W.

Mesoporous carbon-silica composite from low-cost vinasse for the effective adsorption of methyl orange and methylene blue dyes from aqueous systems

Mon, 17 Aug 2026 00:00:00 +0700

The adsorption behaviour of anionic and cationic dyes, namely methyl orange (MO) and methylene blue (MB), respectively, has been studied on carbon-silica composites. These composites were produced through the valorisation of waste vinasse from industrial sugar production as the carbon feedstock, while the silica production was achieved via a simple sol-gel method using low-cost sodium and potassium silicate solutions (Na₂SiO₃ and K₂SiO₃) or higher-cost tetraethyl orthosilicate (TEOS). Use of Na₂SiO₃, K₂SiO₃ and TEOS, led to composites with specific surface areas of 375, 343 and 656 m²/g and % mesoporosity of 79.45, 94.35 and 82.93%, respectively. Adsorption of MO is reduced at higher pH, but MB removal was enhanced under basic pH. Kinetic results demonstrated a strong correlation with the pseudo-second order model, and it was noted that intraparticle diffusion had a key role to play in the mechanism of mass transport. Adsorption data fitted the Sips isotherm model, with excellent adsorption capacity of MO (844.63 mg/g for Na₂SiO₃, 979.06 mg/g for K₂SiO₃, 1161.65 mg/g for TEOS) and MB (514.30 mg/g for Na₂SiO₃, 385.34 mg/g for K₂SiO₃, 294.88 mg/g for TEOS). Elevated temperatures led to preferential removal of the acidic and basic dyes, suggesting an endothermic adsorption process. Carbon-silica composites derived from vinasse are promising materials for the remediation of MB and MO from aqueous systems, exhibiting comparable or higher surface area and adsorption capacities for both dyes relative to other reported materials, while being synthesized through simple preparation processes from low-cost agricultural by-products.

ANN-Based sensitivity analysis for the prediction of energy consumption and head rice yield in a pneumatic paddy dryer

Thu, 15 Oct 2026 00:00:00 +0700

In this study, the performance of a pneumatic paddy dryer was investigated through a sensitivity analysis with an artificial neural network, in order to identify the relative importance of the process variables affecting the energy consumption and head rice yield (HRY). Experiments were conducted using Thai Jasmine 105 paddy rice, which was dried from 24% to 14% wet basis at drying air temperatures of 60°C, 80°C, and 100°C. Increasing the drying air temperature increased the drying rate from 0.46 to 1.11 kg_water/h. Drying at 60°C resulted in the lowest specific energy consumption (SEC) of 3.17 MJ/kg_water and the highest HRY (51.91%), but approximately 6 h was needed to reach the target moisture content for a batch of 20 kg. Higher drying temperatures substantially shortened the drying time but increased the SEC and reduced the HRY to 39.01–41.15%, suggesting a trade-off between drying rate, energy use, and milling quality. The model accurately predicted the moisture content, electrical energy consumption, and HRY, with coefficients of determination (R²) ≥ 0.997 for both the training and testing datasets. Through a sensitivity analysis, the paddy grain temperature (T_paddy) was identified as the most informative predictor, accounting for approximately 38% of the total relative importance, followed by drying time (32%) and drying air temperature (21%). These findings indicate that T_paddy provides valuable information on the transient thermal state of the grain, and can serve as a key process variable for forecasting drying behavior. Rather than representing a standalone commercial drying solution, the framework presented here provides a basis for developing multi-stage drying strategies in which rapid pneumatic pre-drying can be followed by tempering or drying at lower temperatures to strike a balance between drying efficiency and HRY.

Deep learning–based multispectral seed classification for detecting adulteration in KDML-105 rice

Fri, 25 Sep 2026 00:00:00 +0700

This study aimed to develop and evaluate Deep Learning models for detecting varietal contamination in KDML-105 jasmine rice seeds using multispectral imaging and seed morphological features. Four rice varieties—KDML-105 and three Non–KDML-105 varieties (Chainat-1, RD-15, and RD-6)—were used, with 1,000 seeds collected per variety (4,000 seeds in total). All samples were subjected to imaging, object segmentation, and statistical feature extraction across multiple spectral bands to create inputs for three Deep Learning models: a Bayesian Neural Network (BNN), a Convolutional Neural Network (CNN), and a Recurrent Neural Network (RNN). Experimental results showed that the BNN achieved the highest accuracy on multispectral datasets, reaching 0.99 for both the Physical Reflectance and Multi-Reflectance datasets. The CNN produced slightly lower accuracy, while the RNN consistently performed the poorest across all datasets, indicating that feed-forward architectures are better suited for non-sequential reflectance data. SHAP-based feature analysis identified Min_nir as the most influential variable for classifying both KDML-105 and Non–KDML-105 seeds, followed by Mean_blue and several statistical descriptors from the green and blue spectral bands. These findings highlight the strong discriminative capability of NIR reflectance for separating rice varieties with subtle morphological differences, supported by additional surface-related cues captured in the blue band. Overall, the results demonstrate that Deep Learning combined with multispectral imaging is a viable approach for detecting seed contamination. The identification of key spectral bands suggests the potential for designing a lower-cost inspection system using only NIR and blue channels, enabling the development of automated or real-time sorting solutions suitable for practical implementation in Thailand’s rice industry.

FAST Keypoint detection on excess green images for navigation line extraction in maize fields

Niramon Ruangpayoongsak, Thanpisit Rangdeang — Wed, 26 Aug 2026 00:00:00 +0700

Maize field interrow navigation requires high precision for mobile robot heading control because of the narrow soil area between maize rows. This paper presents a new method for extracting a navigation line in the interrow space of maize fields of an automatic granule fertilizer dispenser (AGFD) robot. The proposed algorithm is a crop row detection approach named FAST keypoint detection, which is applied to the excess green image. The FAST keypoints detected are distributed throughout the crop rows in the image. The crop row lines are fitted by the least squares method. Afterward, the navigation line is extracted. The initial region of interest (ROI) is automatically identified using YOLO-based soil segmentation and the adaptive boundary lines are generated. The RMSE and R² value of the navigation lines obtained from the proposed method is compared with that of the excess red- and excess green-based methods. The proposed method achieved a minimum RMSE of 10.06 pixels and a maximum R² value of 0.75. The results reveal that the FAST keypoint detection method on the excess green image outperforms the benchmarks methods for various stages of maize growth with respect to both higher accuracy and less computation time. The proposed method achieved a maximum RMSE improvement of 10 pixels (43.74%), reducing the error from 22.86 pixels, while requiring a computational time of only 53 ms, corresponding to 17% of that of the benchmark method. The results of the proposed method under various field conditions, including differences in illumination, soil color, and vegetation density, are also presented. The proposed method is capable of real-time implementation on the AGFD robot.

Software selection and requirement specification for technology licensing fee management system using AHP and QFD approach

Tinnakorn Phongthiya, Apasara Klaynarong, Chompoonoot Kasemset — Tue, 18 Aug 2026 00:00:00 +0700

Technology licensing fee management in university science and technology parks faces critical challenges including data loss, inefficient royalty tracking, and revenue leakage from spreadsheet-based systems lacking automated notifications and centralized data management. This study develops an integrated framework combining Analytic Hierarchy Process (AHP) with Quality Function Deployment (QFD) for systematic software selection and requirement specification in technology licensing contexts. Conducted at a science and technology park, Chiang Mai University in Thailand, five technology transfer experts validated fourteen initial selection criteria through Index of Item-Objective Congruence (IOC), retaining nine criteria for evaluating three alternatives through AHP pairwise comparisons: customized Booked ME solution, commercial CRM system, and Google Workspace integration. Subsequently, seven end-users and three technical consultants conducted QFD analysis to translate fourteen validated user requirements into twelve technical specifications via House of Quality framework. AHP analysis prioritized security (0.21) and initial cost (0.20) as dominant factors, identifying the customized Booked ME solution as optimal (0.43) over commercial CRM (0.28) and Google-based solutions (0.29). QFD analysis revealed validated calculation algorithms (14.38%), web-based architecture (13.92%), and workflow simplification (11.88%) as highest-priority technical characteristics. The research contributes by: (1) extending AHP-QFD to technology licensing fee management, a novel domain combining financial, legal, and IP management; (2) providing empirical evidence from an emerging economy that platform customization outperforms commercial solutions when security and cost dominate; and (3) developing a replicable integrated framework bridging strategic platform selection with tactical requirement specification.

Hybrid RSM-CCD and ANN-GA optimization of mechanical properties in permanent-backing-plate-assisted GMAW dissimilar welding of AISI 304L and SS400 steels

Nirun Promkesa, Teephet Chaiyason, Suriya Prasomthong — Thu, 13 Aug 2026 00:00:00 +0700

Dissimilar welding between AISI 304L stainless steel and SS400 carbon steel is widely applied in structural and industrial systems, yet differences in thermal and metallurgical properties often lead to heterogeneous microstructures and unstable mechanical performance. This study investigates the mechanical behavior and fracture characteristics of dissimilar AISI 304L–SS400 joints produced by Gas Metal Arc Welding (GMAW) with a permanent backing plate using a hybrid RSM–CCD + ANN–GA modeling and optimization framework. Welding current, arc voltage, and travel speed were selected as the key process parameters and optimized with respect to tensile strength, yield strength, and elongation. Statistical analysis revealed that welding current is the dominant factor controlling mechanical responses through its influence on heat input and weld penetration. The permanent SS400 backing plate introduces a tri-material heat conduction path (304L–SS400–SS400), modifying heat flow and thermal gradients within the weld region. Although the RSM–CCD quadratic models captured the general response trends, the Artificial Neural Network (ANN) provided superior predictive accuracy with lower RMSE, MAE, and MAPE values. Integration of ANN with a Genetic Algorithm enabled efficient multi-response optimization, identifying an optimal welding condition of approximately 170 A, 22.5 V, and 170 mm/min, which was experimentally validated. SEM fractography confirmed ductile micro-void coalescence under the optimized condition and mixed-mode fracture under detrimental parameters. The proposed hybrid framework provides a reliable approach for predicting and optimizing welding parameters in dissimilar GMAW systems with permanent backing plates.

Fuzzy multimodal vehicle routing for last-mile flood relief: A scenario-based model for Chiang Mai

Kanyarat Phutthanawong, Takashi Irohara, Israt Jahan Hridi, Chawis Boonmee — Thu, 27 Aug 2026 00:00:00 +0700

This research addresses the logistical challenges of last-mile relief delivery during flood disasters in Chiang Mai, Thailand. We propose a scenario-based fuzzy multi-depot, multi-trip vehicle routing problem model with split deliveries and integrated multimodal accessibility across varying flood severity levels. The model incorporates fuzzy logic to handle uncertainty by transforming it into an equivalent auxiliary crisp model through triangular membership functions. The objective function aims to minimize total transportation time across five flood scenarios. Results demonstrate that community accessibility requirements vary significantly with flood severity: low water levels can be served entirely by trucks, while severe flooding necessitates intensive boat operations, significantly increasing response times and trip frequencies. Sensitivity analysis reveals that strategic boat-drone collaboration in Scenario 4 can reduce the number of trips by 2 and travel time by up to 3.7%, while significantly simplifying boat routing complexity. The study provides actionable insights for disaster management agencies, emphasizing the critical role of scenario-based planning, multimodal fleet deployment, and robust decision-making under uncertainty.

Geotechnical and environmental evaluation of cement-stabilized lateritic soil with wood bottom ash for road subbase

Sommart Swasdi, Arsit Iyaruk, Thaweesak Thongkhwan, Arun Lukjan — Wed, 19 Aug 2026 00:00:00 +0700

This article investigates the characteristics of the marginal lateritic soil stabilized with cement and wood bottom ash (WBA) as subbase materials. A comprehensive methodology involved modified Proctor compaction and unconfined compression tests, supplemented by microstructural analyses, as well as pilot-scale field construction and leachate monitoring. Results revealed that compressive strength increased with WBA content up to 60%, and was further enhanced by higher cement dosages. Compressive strength was inversely correlated with porosity and voids-to-cement ratio. Field monitoring showed that Cr was the only metal consistently detected in leachate, decreasing over time but remaining above permissible limits, indicating insufficient immobilization due to weaker incorporation into cementitious phases. This study highlights the synergistic effects of WBA and cement in improving soil subbase performance while supporting sustainable reuse of industrial by-products.

Banana peel-derived carbon for supercapacitor electrodes: Carbonization temperature-dependent structural and electrochemical properties

Thu, 13 Aug 2026 00:00:00 +0700

Banana peels (Musa ABB, Namwa cultivar) were utilized as sustainable precursors for the production of carbonaceous materials for supercapacitor electrodes. Carbonization was performed at temperatures ranging from 500 to 1000 °C. During carbonization, CO₂ and H₂O were released, promoting physical self-activation and pore development. X-ray diffraction confirmed the presence of KCl and K₂CO₃ phases, which were removed after washing with HCl and deionized water, leaving only amorphous carbon. Iodine adsorption analysis revealed that increasing carbonization temperature enhanced porosity, with iodine adsorption values of -10.42, 194.20, 150.04, 485.80, 471.44, and 325.09 mg·g^-1 at 500, 600, 700, 800, 900, and 1000 °C, respectively. This trend is in agreement with the change in surface area with carbonization temperature, which increased from 500 to 800 °C (382.403–800.846 m²·g^-1) and decreased at 900 and 1000 °C (684.595 and 156.881 m²·g^-1, respectively). The carbon obtained at 800 °C (CC800) delivered the highest specific capacitance, 169 F·g^-1 at 0.5 A·g^-1 in a 1 M H₂SO₄ electrolyte, owing to its highest specific surface area (800.846 m²·g^-1). Additionally, the redox peaks associated with surface functional groups formed during carbonization would further enhance the capacitance of all carbons. These results demonstrate that banana peel-derived carbon, synthesized via carbonization, exhibits a high surface area and excellent charge storage capability.

Impact of co-existing ions on arsenic remediation using green nano zero-valent iron supported on nanocellulose: A fractional factorial design study

Ratthiwa Deewan, Dickson Yuk-Shing Yan, Pummarin Khamdahsag, Visanu Tanboonchuy — Tue, 30 Jun 2026 00:00:00 +0700

This study focused on developing a novel adsorbent material of green nano zero-valent iron supported by nanocellulose (G-NZVI/NCC) from agricultural waste for the effective removal of arsenic. The adsorption isotherm for arsenite (As³⁺) and arsenate (As⁵⁺) on G-NZVI/NCC showed maximum capacities of 4.2123 and 4.9579 mg·g^‒1, respectively. The model indicated that As³⁺ forms a multilayer on a heterogeneous surface, while As⁵⁺ forms a monolayer on a homogeneous surface. The Gibbs free energy for the adsorption of As³⁺ and As⁵⁺ on G-NZVI/NCC from water demonstrates a non-spontaneous process at higher temperatures for As³⁺, but it occurs spontaneously for As⁵⁺. Furthermore, most previous studies have largely overlooked the influence of co-existing ions in water, or have investigated them individually. Consequently, this study aimed to simulate real-world conditions by examining the simultaneous effects of all relevant ions. This comprehensive approach is critical for the practical implementation of the system, and statistical principles are employed for data analysis and interpretation. However, surface water contains competing ions such as calcium (Ca²⁺), phosphate (PO₄^3‒), bicarbonate (HCO₃^‒), chloride (Cl^‒), and sulfate (SO₄^2‒) that can interfere with arsenic adsorption. A 2^5‒2 fractional factorial design was used to systematically evaluate individual factors, significant parameters, and their interactions on arsenic removal efficiency. The results revealed that Ca²⁺, PO₄^3‒, and HCO₃^‒ play crucial roles in arsenic removal, with Ca²⁺ and PO₄^3‒ enhancing removal efficiency, while HCO₃^‒ exhibits inhibitory effects. Notably, high PO₄^3‒ concentrations unexpectedly enhanced arsenic removal compared to previous studies, attributed to the synergistic physical and chemical adsorption properties of iron oxide and hydroxide phases. Mechanistic analysis revealed that As⁵⁺ can substitute PO₄^3‒ in vivianite formation, creating a vivianite-symplesite solid solution (Fe₃(PO₄)_2.7(AsO₄)_0.3·8H₂O), providing new insights into arsenic immobilization mechanisms under realistic water chemistry conditions.

Mechanical and hydraulic performance of porous asphalt concrete containing recycled PET plastic granule for permeable pavements

Fri, 21 Aug 2026 00:00:00 +0700

This research examines the mechanical and hydraulic properties of porous asphalt concrete (PAC) modified with recycled polyethylene terephthalate granules (G-PET) to enhance the performance of permeable pavement materials. Marshall specimens of the mixture containing various percentages of G-PET of 0%, 5%, 10%, 15%, 20%, and 40% (by weight of asphalt binder) were subjected to Marshall stability, indirect tensile strength, ultrasonic pulse velocity, Cantabro loss, draindown, moisture damage, and permeability tests. The experimental results show that adding G-PET improves pore uniformity, thereby increasing permeability. However, excessive G-PET addition led to higher porosity and reduced stiffness due to particle agglomeration and weak interfacial bonding of the mixture. Quantitative results showed that the 10% G-PET mixture provided the most favorable balance, meeting volumetric, mechanical, and permeability criteria within functional specifications for permeable pavements. This finding fills a gap in sustainable pavement design by developing an experimentally validated method to incorporate recycled plastics into porous asphalt mixes, thereby supporting circular-economy goals for infrastructure materials.

Mechanical and durability performance of hybrid nano clay–CNT concrete: Experimental investigation and data-driven polynomial regression modeling

Wed, 23 Sep 2026 00:00:00 +0700

Concrete remains the most widely used construction material, yet its brittleness, microcracking susceptibility, and durability limitations continue to challenge the development of long-lasting infrastructure. At the same time, reducing the environmental impacts associated with cement production requires innovative materials that improve concrete performance. Nanomaterials such as nano clay and carbon nanotubes (CNTs) offer potential for modifying the cementitious matrix and enhancing mechanical and durability properties. However, their combined effects, particularly when integrated with predictive modelling, remain insufficiently investigated. This study evaluates M35 concrete incorporating 0–6% nano clay as a partial replacement of Ordinary Portland Cement and 0–0.20% multi-walled carbon nanotubes (MWCNTs). Compressive, flexural, and split tensile strengths were evaluated at 7, 14, and 28 days, while water absorption was considered as a durability indicator. The optimum hybrid mixture containing 4% nano clay and 0.10% CNTs exhibited improved mechanical performance compared with the control mixture. Second-degree polynomial regression models were developed using nano clay content, CNT dosage, and curing age as predictor variables. The models demonstrated high predictive capability, with R² values ranging from 0.94 to 0.99. Sensitivity analysis identified nano clay content, CNT dosage, and curing age as important parameters influencing concrete performance. The combined incorporation of nano clay and CNTs improved strength and reduced water absorption, indicating enhanced matrix densification and resistance to crack initiation and propagation. The proposed experimental and data-driven framework provides a useful approach for predicting the performance of hybrid nanomaterial-modified concrete and supports the development of durable and sustainable concrete for long-term infrastructure applications.

Drone-based application of inflorescence-inducing chemicals in longan orchards: Effects of flight altitude and mixing ratios

Tue, 23 Jun 2026 00:00:00 +0700

Drone-assisted chemical application in longan orchards is feasible but requires refined methods for efficient and uniform aerosol delivery. This research aimed to determine the optimal drone flight altitude and chemical-to-water mixing ratio for inducing inflorescence in longan trees. Three chemical mixtures were tested: the conventional farmer-used ratio (50 g/10 L), a higher concentration (50 g/5 L), and a reduced dosage (25 g/5 L). Droplet deposition at different flight altitudes above the longan canopy was evaluated using water-sensitive paper to identify the most effective spraying conditions. The results revealed that at a drone flight altitude of 2 m above the longan canopy, the droplet deposition percentage from the drone sprayer reached a maximum of 76.22% on water-sensitive paper. This flight altitude was selected for the flight test to determine the appropriate mixing ratio of inflorescence-inducing chemicals for spraying in longan orchards. The reduced dosage treatment (25 g/5 L) resulted in the greatest number of inflorescences and the highest yield. Although the conventional and reduced treatments share the same chemical concentration (5 g/L), the latter involved a lower total amount of chemical applied per tree due to the reduced spray volume during drone operation. This approach enables farmers to decrease total chemical usage and operational costs while enhancing inflorescence induction and yield in longan orchards. Additionally, it enables faster drone operation, effectively doubling the treatment area per flight.

AI-assisted power quality improvement in PV systems using RNN-based DSTATCOM and HiClampCasBoost converter

Fri, 31 Jul 2026 00:00:00 +0700

The rapid deployment of Photovoltaic (PV) systems in modern power distribution networks has caused major issues related to voltage fluctuations, reactive power unbalance, harmonics, and less efficient systems. It is critical to protect Power Quality (PQ) under these conditions to ensure safe and consistent operation of grid. Flexible AC Transmission System (FACTS) devices are used to mitigate those problems encountered in the network. Hence, this work proposed a grid-tied PV system integrated with Distribution Static Compensator (D-STATCOM) device that help accurate voltage sag, swell, and harmonic distortion at the distribution level. In this work, the interaction of D-STATCOM with the grid is controlled by a Recurrent Neural Network (RNN) based controller with PWM signals adjusted for voltage stability, sag, swell and load disturbances compensation. Further, a high-gain HiClampCasBoost Converter (HiClampCasBoost) converts low PV output voltage and regulates the DC-link, assuring constant output of the D-STATCOM while under voltage conditions prevail. Additionally, a Adaptive Neuro-Fuzzy Inference System (ANFIS) based Maximum Power Point Tracking (MPPT) controller, optimized using the Hummingbird Optimization Algorithm (HOA) for tracking Maximum Power Point (MPP) of PV array under dynamic solar conditions to produce maximum energy from PV array whilst ensuring D-STATCOM DC-link stability. The MATLAB simulations show that the developed system enhances PQ and grid stability in comparison to conventional methods. The proposed system has several benefits, such as superior voltage regulation, reduced THD of 1.63%, superior converter efficiency of 96%, tracking efficiency of 99% superior dynamic operation performance, etc.

Optimizing layering configurations for enhanced flexural and impact performance in perforated Al foil/glass fiber/epoxy hybrid laminates

Tue, 30 Jun 2026 00:00:00 +0700

Lightweight panels for aircraft, vehicles, and protective structures need to absorb damage without a large weight penalty. In practice, many fiber metal laminates (FMLs) still struggle with shaping or curvature and can delaminate early, largely because the metal plies are relatively thick and the metal/polymer interface is prone to separation. An alternative hybrid laminate was developed using thin perforated Al foil (A) (~0.1 mm; perforations ~150 μm) combined with woven glass fiber (G)/epoxy, and the influence of stacking sequence on flexural and impact behavior was examined. Five configurations were manufactured by hand lay-up and evaluated using three-point bending (ASTM D790) and unnotched Charpy impact (ASTM D6110). Fracture surfaces and interfacial features were examined by optical microscopy and SEM. The laminates showed a clear, design-driven performance balance between bending strength and impact resistance. The symmetric, staggered lay-up V4 (GGGAAAGAAAGGG) delivered the highest flexural strength (229.74 MPa), whereas the metal-rich core V3 (GGGAAAAAAGGGG) produced the highest impact resistance (123 kJ/m²). Microstructural characterization indicates that resin flow through the perforations forms mechanical “anchors” at the Al foil/epoxy interface, limiting adhesive-type separation and promoting more progressive damage through cohesive cracking and controlled delamination. The results suggest that perforated Al foil/glass/epoxy hybrids offer a practical route to tune stiffness and toughness for lightweight skins/panels and protective components requiring high bending performance with moderate-to-high impact resistance.

Forecasting dengue severity using machine learning and environmental predictors in Chanthaburi, Thailand

Tue, 24 Mar 2026 00:00:00 +0700

Dengue fever is a vector-borne disease whose dynamics are substantially driven by climate change and environmental variability. To this end, this study proposed the establishment and assessment of machine learning models for predicting the severity of dengue incidents in Chanthaburi province, Thailand, based on climatic and environmental variables from 2018–2022. Both classification accuracy and confusion matrix analyses were used to compare the implementation of four machine learning algorithms: Random Forest, Gradient Boosting, Extra Trees, and CatBoost. The highest-performing model was able to predict a sum score of 0.50 correctly 86.70% of the time, which suggests that the developed system has good predictive ability, especially in homing in on low-severity dengue cases. Still, challenges associated with recognizing high-severity cases persist. Shapley Additive Explanations (SHAP) sensitivity analysis revealed that specific air pollutant levels (PM10, PM2.5), time-lag parameters, and indicators such as temperature and humidity, particularly during certain periods, were significant predictors of dengue severity. Our results reveal the intricate relationships between environmental variables and the pattern of dengue transmission, arguing for a judicious use of machine learning tools as evidence-based support to inform disease control policies. Future studies that consider other variables, longer time series data, and advanced modeling techniques are needed to increase the accuracy of predictions, especially with respect to improving sensitivity for high-severity dengue outbreaks.

Lightweight IOT-friendly image encryption scheme based on Latin square and dynamic DNA coding

Sameeh Abdulghafour Jassim, Seddiq Q. Abd Al-Rahman, Ahmed Adil Nafea — Wed, 29 Apr 2026 00:00:00 +0700

The fast growth of Internet of Things (IoTs) devices calls for the demand to accomplish an efficient and lightweight image encryption. Also, protecting the privacy of sensitive visual information on devices with limited resources presents significant challenges. Existing techniques and methods are not able to provide both powerful security and low computational requirements. Therefore, we propose an innovative encrypting scheme that combines three methods (Latin squares, dynamic DNA sequencing, and chaotic systems). This architecture is designed for the IoT environments where it is important that transactions are made both efficient and secure. Additionally, our approach offers strong security assurances and suitable for the constraints of edge devices. The performance evaluations of the proposed technique against modern encryption algorithms display a promising result. Our proposal demonstrates almost perfect randomness. It achieved an image entropy of (7.9994) and a minimal correlation of (-0.00472). The proposed system also demonstrated its high ability to resilience the differential attacks, with NPCR and UACI of (98.671) and (34.098), respectively. Also, the keyspace of the proposed method is ( ), that providing a robust defense against brute-force attacks. Despite this high security level, the technique maintains practical performance, encrypting a (512×512) image in (1.4026) seconds. These outcomes confirm the algorithm's effectiveness for protecting sensitive images in IoT applications, offering both security and suitable efficiency for real-world use.

A structured maturity model for evaluating smart logistics parks transformation: Case study in western China

Huiqiong Huang, Woramol Chaowarat Watanabe — Tue, 30 Jun 2026 00:00:00 +0700

The transformation of traditional into smart logistics parks (SLPs) requires a structured framework for assessing maturity and guiding development. Existing maturity models, often adapted from smart city or industrial park frameworks, lack specificity and empirical validation for the logistics park context. To address this gap, this study constructs and validates a data-driven maturity model for SLPs by integrating a literature review, expert consultation, and case studies of two logistics parks in western China. The model comprises five dimensions—Smart Economy, Public Services and Smart Governance, Infrastructure and Intelligent Technology Application, Skilled Human Capital, and Environmental Sustainability—and defines 20 measurable factors. These factors are evaluated through a four-level maturity framework progressing from Initial to Optimization stages, corresponding to the conventional levels of Initial, Defined, Managed, and Leading. The application revealed that both parks were positioned between the Initial and Growth stages. Logistics Park 2 excelled in digital infrastructure and technology adoption, as well as workforce digital literacy, whereas Logistics Park 1 enjoyed advantages in operational ROI and stakeholder integration. While for both parks, Public Services and Smart Governance reached the Growth stage, Smart Economy and Environmental Sustainability lagged, reflecting reliance on volume-based growth and underdeveloped sustainability outcomes. The model differentiated between transitional and emerging SLPs, proving its value as a diagnostic and strategic tool. One limitation, however, is the model's initial validation within the two case studies in western China, and future research should expand the application to diverse contexts to enhance generalizability and robustness.

Integrating low-cost hydroacoustic surveying for canal sedimentation monitoring: a case study from southern Thailand

Wed, 17 Jun 2026 00:00:00 +0700

Efficient sedimentation monitoring is vital for maintaining navigability and hydraulic performance in canal systems, yet conventional hydrographic surveys remain costly and logistically demanding. This study integrates a low-cost hydroacoustic surveying method using a recreational-grade single-beam echosounder to provide cost-effective and practical solution for canal sedimentation management. The approach was first validated in Songkhla Lake, representing mixed freshwater, brackish, and saline conditions, and subsequently applied to the Samrong Canal in southern Thailand as a real-world case study. Controlled experiments compared echosounder-derived depths with reference measurements to evaluate vertical accuracy under different water types following the International Hydrographic Organization (IHO) S-44 Special Order criteria. Results showed a standard deviation (SD) of 0.05 m and RMSE₉₅ = 0.14 m in controlled settings, with no statistically significant effect of water type on accuracy. Semi-controlled field surveys yielded SD values of ±0.06 m (fresh), ±0.05 m (brackish), and ±0.09 m (saline), all within the IHO tolerance limits. The validated setup was then applied for 0.5-m contour bathymetric mapping in Samrong Canal, successfully delineating shoaling zones and siltation hotspots critical for maintenance planning. The findings demonstrate that integrating low-cost echosounders into canal monitoring workflows can produce IHO compliant, high-utility bathymetric data, offering a scalable and cost-effective alternative for sedimentation assessment in shallow tropical waterways.

Feature-level fusion of Sentinel-1 SAR and Sentinel-2 optical imagery for crop-type mapping in humid tropical mountainous terrain

Indarto Indarto, Gunawan Gunawan, Mahrus Irsyam, Fahmi Arif Kurnanto — Mon, 12 Oct 2026 00:00:00 +0700

Accurate crop-type mapping in humid tropical mountains is challenged by persistent cloud cover, rugged topography, and fragmented smallholder parcels. This study evaluates a feature-level fusion workflow that integrates Sentinel-2 optical indices and Sentinel-1 SAR backscatter for crop-type mapping in Candipuro District, Lumajang, East Java, Indonesia. Sentinel-2 surface reflectance and Sentinel-1 IW data acquired in June–July 2023 were harmonised in Google Earth Engine to a common 10 m pixel grid, so that all comparison maps were evaluated on the same spatial support. The optical component comprised NDVI, NDBI, and NDWI, whereas the radar component used the VH/VV polarisation ratio. A DEM was used only to characterise the mountainous setting and to assist terrain interpretation, and was not included as a predictor in the classifier. The stacked feature image was classified using a Random Forest model trained with 101 field-survey ground control points representing eight land-cover/crop-related classes. Accuracy was assessed using 393 stratified random validation points. The fused-image classification achieved an overall accuracy of 93% and a kappa coefficient of 0.90. The fusion workflow yielded richer thematic differentiation than the optical-only comparison scenarios, particularly for water bodies, sugarcane, and shrubland. However, class-level results reveal substantial uncertainty for corn fields and moderate confusion in built-up and water classes. Therefore, the present workflow should be interpreted as a site-specific, single-date proof of concept rather than a fully transferable operational solution. The results indicate that Sentinel-1/Sentinel-2 feature fusion is promising for crop-type mapping in cloud-prone mountainous landscapes, while future work should incorporate multi-temporal phenology, richer vegetation indices, SAR texture features, uncertainty mapping, and broader field sampling.