RMUTL Engineering Journal

Machine Learning Approaches for Malaria Forecasting Using Environmental Drivers: A Case Study in Tak Province, Thailand

2025-12-12T14:43:58+07:00

Malaria remains a significant public health challenge in Thailand's border provinces, were traditional reactive surveillance limits outbreak prevention capabilities. This study systematically evaluated six machine learning algorithms (Support Vector Machine (SVM), K-Nearest Neighbors (KNN), Ridge regression, Elastic Net, Lasso, and XGBoost) for operational malaria forecasting at 1-4-week horizons in Tak Province, Thailand. Using 13 years of surveillance data (2012-2024, n=689 epidemiological weeks) integrated with satellite-derived environmental predictors (rainfall, temperature, soil moisture) processed via Google Earth Engine (GEE), models were trained using chronological partitioning and evaluated on 2024 holdout data using coefficient of determination (R²) and root mean square error (RMSE). Algorithm-specific optimal performance was identified: SVM achieved superior 1-2-week forecasting (R² = 0.744 and 0.687, RMSE = 14.6 and 16.2 cases/week), while KNN excelled at 3-4-week horizons (R² = 0.748 and 0.731, RMSE = 14.7 and 15.2 cases/week). Statistical significance testing with bootstrap confidence intervals confirmed genuine algorithmic advantages rather than random variation. Historical case features dominated predictive performance, while environmental variables provided complementary information. Models successfully tracked temporal patterns including the 2022-2023 transmission rebound. The satellite-based framework provides scalable solutions for resource-limited settings, with 1-4-week lead times enabling proactive intervention planning to support Thailand's malaria elimination objectives. This operational forecasting approach offers a replicable template for similar endemic contexts across Southeast Asia.

Optimization of Vertical Screw Conveyor for Biomass Sampling: Influence of Pitch Geometry on Energy Dissipation and Material Integrity

2026-02-04T11:12:57+07:00

This study investigates the optimization of vertical screw conveyor design to enhance the efficiency and reliability of automated cassava sampling systems in agro-industrial processing. A comparative analysis was conducted between two screw configurations: Type B (P/D = 0.67) and Type C (P/D = 0.50), with varying screw diameters (0.11-0.20 m) and rotational speeds (56.78, 48.67, and 36.50 rpm for 18-, 21-, and 28-tooth gear sets, respectively) under controlled hydraulic operation (p = 152 bar, Q_oil = 4.875×10⁻⁴ m³/s, P_in = 7,410 W). The results indicate that over 99% of the hydraulic input power is theoretically dissipated as heat, based on the first-law energy balance. The Type B configuration, particularly at a diameter of 0.20 m and maximum rotational speed (56.78 rpm), demonstrated superior performance, achieving the lowest Specific Energy Consumption (SEC) of 23.25 kJ/kg, representing a 25.0% reduction in SEC and a 33.4% gain in useful mechanical work output (P_out: 14.07 vs. 10.55 W) over Type C. Furthermore, the wider pitch of Type B effectively mitigates material compaction and reduces cumulative frictional stress during transport, preserving the physical integrity of cassava chips (bulk density: 496.4 ± 62.4 kg/m³; moisture content: 13.5 ± 0.5% w.b.). The study concludes that the Type B configuration is the optimal design for maximizing flow stability and minimizing energy loss while satisfying quality assurance requirements.

Optimization of Swine Farm Wastewater Treatment Using Mixed Microalgae: Statistical Modeling and Performance Evaluation via Response Surface Methodology

2026-01-07T14:54:39+07:00

Swine farm wastewater contains high concentrations of organic matter and nutrients, requiring treatment approaches that are both effective and economically feasible. This study evaluated the performance of mixed indigenous microalgae for the removal of chemical oxygen demand (COD) and total phosphorus (TP) from swine farm wastewater and optimized the effects of initial pH and algal concentration using Response Surface Methodology (RSM) with a Central Composite Design (CCD). Thirteen experimental runs were conducted under outdoor conditions to reflect field applicability. COD and TP removal efficiencies ranged from 74.30–83.85% and 75.17–83.58%, respectively. Statistical analysis showed that pH significantly influenced both COD and TP removal, whereas algal concentration exerted a stronger effect on COD removal but a comparatively weak influence on TP, in agreement with ANOVA results. The quadratic model demonstrated strong predictive performance for COD (R² = 0.9879; predicted R² = 0.9245), while the TP model displayed limited predictive capability, suggesting that additional unmeasured factors may govern phosphorus reduction. Numerical optimization identified pH 7.69 and algal concentration (A600) 2.072 as the optimal conditions, yielding predicted removals of 78.21% COD and 82.00% TP. Although TP levels approached regulatory thresholds, COD remained above discharge limits, highlighting the need for a polishing step prior to release. Overall, the results demonstrate that mixed indigenous microalgae offer a robust, low-cost treatment strategy for swine wastewater and provide optimized operational conditions to support practical implementation.

Enhancing Gypsum Ceiling Sheets with Malt Waste: Optimal Composition for Strength and Insulation

2026-03-31T09:37:21+07:00

This study investigates the use of malt waste, a byproduct of the beer production process, as a bio-based additive in gypsum ceiling sheets to enhance mechanical performance while improving thermal insulation properties. Gypsum composite specimens were prepared by incorporating malt waste at weight ratios of 100:0, 90:10, 85:15, 80:20, 75:25, and 70:30, and the effects on density, thermal conductivity, and bending strength were systematically evaluated and compared with those of conventional gypsum boards complying with TIS 219-2009. The results show that increasing malt waste content led to a reduction in density from 1.16 g/cm³ for pure gypsum to 0.76 g/cm³ at a 70:30 ratio, representing a decrease of approximately 34.5%. Similarly, thermal conductivity decreased to a minimum value of 0.094 W/m·K, indicating improved insulation performance compared with conventional gypsum boards. However, the 85:15 gypsum-to-malt waste ratio demonstrated the optimum balance between thermal insulation and mechanical performance, achieving the highest longitudinal flexural force of 297.57 N and a thermal conductivity of 0.097 W/m·K. These findings highlight the potential of malt waste as a sustainable reinforcement material for the development of eco-friendly gypsum ceiling applications.

Rail Freight Container Management System Utilizing Radio Frequency Identification and Cargo Tracking Technology

2025-11-21T09:25:23+07:00

Real-time monitoring of cargo in transit is essential for improving safety and operational efficiency in public railway transportation systems. Although technologies such as Radio Frequency Identification (RFID) and Global Positioning System (GPS) have been widely applied in logistics, their systematic implementation within the Railway of Thailand remains limited. This paper proposes a real-time railway freight container management system designed to reduce operational errors and prevent cargo loss. The proposed system integrates UHF RFID technology with GPS-based tracking, enabling continuous identification and localization of cargo containers. Tracking data is transmitted to a centralized web-based platform via wireless communication networks for real-time monitoring and management. Experimental evaluation conducted under real operational conditions demonstrates that the system achieves an average end-to-end latency of 185 ms, with a maximum observed delay of approximately 310 ms, confirming its suitability for real-time railway logistics operations. GPS-based container localization yields a mean absolute positioning error of 4.8 meters, with a 95% confidence interval of 4.2–5.4 meters, based on 200 ground-truth samples. Comparative analysis shows that the parcel misplacement rate is reduced from 2.3% under manual handling procedures to 0.4% after system deployment, representing an 82.6% reduction in operational errors. The software platform adopts standard design patterns to enhance reliability and maintainability, while secure authentication is ensured using a bcrypt-based hashing mechanism. The results confirm the practicality and effectiveness of the proposed system for large-scale railway freight logistics.

Sustainable Acoustic Absorbers: Fabrication and Sound Absorption Performance of Compression-Molded Composites Derived from Bamboo Leaf Waste

2026-01-26T12:51:55+07:00

This research addresses the issue by developing a novel, eco-friendly acoustic absorbing material that aligns with the principles of the Bio-Circular-Green (BCG) Economy. The study utilizes dry bamboo leaf powder as the primary natural fiber and Persea kurzii powder, a natural resin/binder, as the binding agent. Standard circular specimens were fabricated using a compression molding technique. Three different ratios of fiber-to-binder were investigated: Sample A (90:10), Sample B (80:20), and Sample C (70:30). The sound absorption properties were strictly tested using the Transfer-function method (ISO 10534-2). The results demonstrated that the specimen with highest fiber loading, Sample A (90:10), exhibited the most effective acoustic performance. This optimal sample achieved a high Noise Reduction Coefficient (NRC) of 0.44 and showed maximum absorption with a Sound Absorption Coefficient (SAC) of 61.96% at a key mid-frequency of 500 Hz. The results strongly suggest that a higher proportion of bamboo leaf fiber is crucial for developing and maintaining the necessary porous structure within the composite, which facilitates effective sound dissipation. These findings demonstrate the potential of bamboo leaf composites as sustainable, cost-effective alternatives to synthetic sound-absorbing panels, successfully valorizing agricultural waste into high-value products.

RFID-Driven Smart Border Passing Architecture for Cloud-Integrated Vehicle Authentication

2025-12-08T11:02:30+07:00

As global trade and tourism continue to expand, cross-border vehicle traffic has significantly increased, demanding more efficient and secure inspection systems. Conventional border checkpoint operations in Thailand rely on manual document verification, which is time-consuming and error-prone. This study proposes the design and implementation of a smart border passing system using Radio Frequency Identification (RFID) to enhance the identification and monitoring of vehicles at border checkpoints. The proposed system integrates three main components: a portable RFID reader, a flexible RFID tag sticker, and a cloud-based database dashboard for data synchronization and tracking. The RFID reader was developed as a compact, battery-powered device capable of reading UHF tags and transmitting data via Wi-Fi in real time. Experimental results demonstrate that the reader achieved 100% detection accuracy within 1.8 meters, with reliable tag readability for 30 days under real outdoor conditions. The web-based dashboard successfully displayed ENTRY, EXIT, and TIMEOUT events, enabling efficient vehicle tracking and administrative reporting. The system offers a practical and cost-effective solution for smart border management, reducing manual workload and supporting proactive detection of unauthorized crossings or vehicle theft. Future enhancements will focus on improving adhesive durability, expanding wireless communication, and integrating predictive analytics for real-time anomaly detection.