Srinakharinwirot University Engineering Journal

Real-Time Federated Learning for Smart Energy Management in an Educational Building at RMUTSV

Santi Karisan, Suporn Rittipuakdee, Santiphong Khongkaeo, Sittisak Rojchaya — Mon, 06 Jul 2026 00:00:00 +0700

Efficient electrical energy management in the Industrial Technician School Building at RMUTSV is crucial due to the complex and continuously fluctuating load characteristics. This research proposes a real-time framework based on Federated Learning (FL) for multi-phase load forecasting (Phases A, B, and C) and adaptive anomaly detection. The system integrates deep learning techniques, including Long Short-Term Memory (LSTM), Gated Recurrent Unit (GRU), and Autoencoder networks. Experiments using real-time time-series data collected over a four-month period (December 2024 - March 2025) revealed that the FL model achieved a Mean Absolute Error (MAE) of 1.24A, comparable to the centralized model (MAE = 1.15A). Phase B exhibited the highest volatility, with 41 anomalies detected (54.7%). A positive correlation between load and temperature was observed (r = +0.68). The total loss graph decreased within 50 training rounds, indicating strong learning capability. After implementation, the system reduced peak load by an average of 6.3% and significantly decreased abnormal events, enhancing stability and minimizing energy loss. This study demonstrates that FL is an effective, secure, and sustainable approach for deploying Artificial Intelligence (AI) in smart grid systems.

Enhancing Engineering Design Skills and Environmental Awarenessthrough a STEM-Based Educational Game

Chinnaphat Ekwaraphaisankul, Nattapat Klaisuwan — Mon, 06 Jul 2026 00:00:00 +0700

Addressing climate change necessitates educational innovations to foster engineering problem-solving skills among youth. This research aims to evaluate the efficiency and effectiveness of a STEM-based educational game in enhancing engineering design skills and environmental awareness. The participants were 36 eleventh-grade students selected via stratified random sampling. Research instruments included the ‘Zero Carbon’ board game, a situational test for engineering design skills, and an environmental awareness questionnaire. Data were analyzed using descriptive statistics and a one-sample t-test against an 80% criterion. The findings revealed that the game innovation met the efficiency criteria (E1/E2 = 80.93/83.05). Moreover, students’ post-intervention engineering design skills and environmental awareness were significantly higher than the 80% criterion (p < .05). This study thus confirms that a STEM-based educational game is an effective instrument for developing engineering competencies and fostering sustainable environmental consciousness.

The Study and Analysis of Causes and Preventive Guidelines for Fire Incidents in Plastic Pellet Factories by What If Analysis

Kanokpath Buajan, Komsan Hongesombut, Parnjit Damrongkulkamjorn — Mon, 06 Jul 2026 00:00:00 +0700

This research aims to investigate the factors contributing to fire incidents and to assess the risk level in a recycled plastic pellet manufacturing plant. Frequent fire occurrences in the past, each causing severe damage, highlight the necessity of conducting a systematic study using the What-If Analysis technique. This flexible qualitative method enables the identification of potential hazards even in the absence of prior incidents and facilitates risk prioritization through the development of a structured risk assessment. The analysis revealed six major risk factors: deteriorated machinery and lack of maintenance, operational errors by workers, unsafe storage of flammable materials, absence of documents for operating procedures, the nonstandard electrical system installation, and insufficient firefighting equipment and emergency-drill readiness. This study applied the What If Analysis technique to assess the risks in the recycled plastic pellet of a case study factory, involving a total of twenty-six events across twenty-one production steps. The results indicated that two events (7.69%) fell under Risk Level 4 (Unacceptable), all associated with the melting process. Additionally, eleven events (42.31%) were classified as Risk Level 3 (High Risk) and were found in shredding, high-speed washing, dewatering, hot air drying, plastic storage, and certain parts of the melting process. Meanwhile, thirteen events (50.00%) were categorized as Risk Level 2 (Acceptable with Control) and occurred in feeding, sorting, conveying, general washing, pellet cutting, and drying steps. The study results can lead to recommendations for developing an effective risk management plan for the factory. The outcomes of this study provide a practical guideline for systematically assessing fire risk for other recycled plastic pellet factories. Moreover, it can be applied to other industries with similar thermal and mechanical processes.

Development of National Freight Transport Analytics System Using GPS-Based Truck Data

Treerapot Siripirote — Mon, 06 Jul 2026 00:00:00 +0700

Road freight transport by trucks is the dominant freight transport mode in Thailand and plays a crucial role in the national economy and supply chains. However, freight transport planning and management at the national level has long been constrained by the lack of detailed data that accurately reflects real freight movement patterns in both spatial and temporal dimensions. This academic article aims to present the development process of National Freight Transport Analytics system using GPS-based truck data. The system utilizes GPS data collected from trucks, integrated with big data analytics techniques, to generate insights into freight movement patterns, supply chain structures, and truck operational behaviors. The developed system (validated with average trip lengths and traffic counts on road within acceptable limit’s results) is capable of processing massive volumes of GPS data, amounting to hundreds of millions of records per day, and visualizing the results through key performance indicators (KPIs). For instance, the system is capable of analyzing the overall freight movements of more than three million trips per month, derived from more than 5,000 of unique activity patterns at truck stop points and commodity trip chain. These outputs support evidence-based policymaking, logistics infrastructure planning, and freight transport management at the national scale.

Development of a Semi-automatic Seasoning Dispersion Control System for Fried Cassava Chips for Ban Khao Kloy Fried Cassava Community Enterprise in Songkhla Province

Sahapong Somwong, Thanwit Naemsai, Worapong Boonchouytan, Athirot Mano — Mon, 27 Jul 2026 00:00:00 +0700

This research aims to develop a control system for coating seasoning dispersion for fried cassava to solve the problem of uneven coating that causes inconsistent product quality. This results in higher costs and reduced revenue for the Ban Khao Kloi Community Enterprise Group in Songkhla Province. The system, designed using food-grade materials in accordance with GMP standards, consists of a rotating drum powered by a 220-volt, 4,500-rpm AC motor and a geared motor with a ratio of 10:1–40:1, controlled by a 3-way selector switch. A high-pressure fan and air ducts are also installed, controlled by a microcontroller to precisely determine the time and frequency. Experimental results showed that the weight of the raw materials and the mixing time affect the thoroughness of the coating. 35 grams of seasoning powder is used for 0.5 kilograms in 5 minutes, and 140 grams for 2 kilograms in 11 minutes. The system can automatically spray and stop working according to the settings. In practice, it can reduce seasoning powder usage by 25–30% and breakage of fried cassava chips by 10–15% at a production volume of 2 kilograms. The results indicate that this system reduces raw material loss and improves product quality. and improve production efficiency, which is a development approach that helps create sustainable income for local communities.

The Development of an Automated System for Air Conditioner On/Off Control Using Remote Signal Emulation

Tanin Duangchan, Siripong Chaysin — Mon, 27 Jul 2026 00:00:00 +0700

Some air conditioners used in buildings or equipment storage rooms do not provide an automatic on/off scheduling function, or their timer functions have operational limitations. As a result, personnel are required to manually control the operation of these air conditioners. This research aimed to develop an automatic air conditioner control system using a microcontroller to emulate the infrared signals transmitted by the original remote control. The proposed system was implemented by employing an ESP8266 development board together with an infrared receiver to capture and record signal data from the air conditioner remote control. The acquired signals were then analyzed and used to generate corresponding infrared command signals for controlling the air conditioner. In addition, a Real-Time Clock (RTC) module was integrated into the system to enable automatic scheduling of on/off operations. The prototype system was tested with an air conditioner installed in an electrical measuring instrument storage room. The operating schedule was configured to automatically turn the air conditioner on from 10:00 a.m. to 4:00 p.m., while room temperature data were continuously recorded over a period of 15 days. The experimental results demonstrated that the developed device successfully transmitted the control signals and operated the air conditioner according to the predefined schedule in every test, achieving a 100% success rate. Furthermore, the recorded temperature data indicated a continuous decrease in room temperature during the periods when the air conditioner was activated. The findings confirm that the proposed system can effectively replace conventional remote-control-based operation of air conditioners. The developed system reduces reliance on human intervention while maintaining reliable operational performance. Moreover, it can be adapted for use with other brands and models of air conditioners that are controlled via infrared remote signals.

Comparative Analysis of LQR Control Performance between Linear and Discrete Time Nonlinear Inverted Pendulum Models

Thosapon Krinsukont, Taweedej Sirithanapipat — Mon, 10 Aug 2026 00:00:00 +0700

This paper presents an analysis and performance evaluation of a Linear Quadratic Regulator (LQR) synthesized from a linearized model and applied to stabilize a cart–inverted pendulum system with nonlinear dynamics in a discrete-time framework. The study focuses on investigating the effects of system nonlinearities on control accuracy and robustness through simulation-based comparisons between linear and nonlinear models at initial pendulum angles of 15° and 30°. The results show that at an initial angle of 15°, the responses of both models are closely aligned. In contrast, at 30°, noticeable discrepancies arise due to nonlinear effects, with maximum deviations of approximately 25%, 18%, and 14% observed in cart position, pendulum angle, and cart velocity, respectively, compared to the linear model. Nevertheless, the LQR controller successfully maintains system stability and drives the system toward the equilibrium point. These findings highlight the practical limitations of linearized control design when applied to highly nonlinear systems and suggest future opportunities for advanced data-driven or artificial intelligence-based controllers to further enhance performance, reduce model mismatch, and improve adaptability under complex operating conditions.

Effect of Using Bottom Ash as Fine Aggregate on Mechanical Properties and Thermal Conductivity of Mortar

Aunchana Kijjanon, Lyna Prak , Taweechai Sumranwanich — Thu, 13 Aug 2026 00:00:00 +0700

This research aims to study the effect of using bottom ash as a fine aggregate on the compressive strength, split tensile strength, porosity, density, thermal conductivity of mortar, and internal temperature of mortar-based structural members exposed to the real environment. Hydraulic cement was used as the principal binder, with fly ash replacing a portion of the hydraulic cement at 30% by weight of the total binder. Bottom ash was used to replace fine aggregates at 0, 10, and 30% by weight of fine aggregate, respectively. All mortar mixtures had a constant water-to-binder ratio of 0.40. From the study results, it was found that the mortar with 30% fly ash and 10% partial fine aggregate replacement with bottom ash (FA30BA10) had the highest compressive and split tensile strength. This was due to the water retention property of bottom ash and the pozzolanic reaction of fly ash, which increased the density and decreased the porosity of the mortar, resulting in increased compressive and split tensile strength. The mortar using 30% fly ash and 30% bottom ash (FA30BA30) had the lowest thermal conductivity. The high amount of bottom ash increased the porosity of the mortar. This made it difficult for heat to transfer through the material, resulting in the lowest thermal conductivity. Furthermore, structural member formed using mortar containing 30% fly ash had the lowest internal temperature. The maximum temperature range for the structural member formed with mortar containing 30% fly ash was 1.91% - 6.63%, which was lower than that of the structural member formed using control mortar.

Optimization of Environmental Control in Smart Silkworm Houses using Computational Fluid Dynamics (CFD) and Multi-Physics Simulation

Whichaya Ringkananont — Mon, 17 Aug 2026 00:00:00 +0700

This research presents the design of the internal environment of a silkworm-rearing greenhouse using computational fluid dynamics (CFD) to optimize temperature, humidity, and ventilation control for tropical climates. A three-dimensional model of a 4 × 6 × 3 meter greenhouse was created and analyzed with OpenFOAM, considering air flow, heat transfer, and moisture distribution under climate conditions representative of Thailand. Simulation results show that the proposed design approach reduces internal temperature variability, minimizes stagnant-air zones, and improves humidity uniformity compared with a conventional greenhouse—reducing cumulative temperature by 8.2%, achieving a 9.1% improvement in moisture distribution uniformity, and yielding a mean air-density difference of 0.9% between simulation and measurements. In addition, appropriate placement of openings and ventilation rates enhanced airflow performance without creating localized strong drafts. Results from this study demonstrate the potential of applying CFD modeling to the design of agricultural buildings that can effectively control the environment and accommodate climate change in tropical regions.