Jurnal E-Komtek (Elektro-Komputer-Teknik)

A Machine Learning-Based Early Warning System for Electricity Outage Due to Extreme Weather

2026-02-03T11:37:04+07:00

Electricity is a critical resource that supports various sectors in Indonesia, especially during extreme weather. Outages have become serious for operational risks during extreme weather. This study proposes a machine learning-based early warning system to predict electricity outages caused by extreme weather. Historical weather and outage data were combined using spatial alignment. Key innovation of this study involved geospatial feature enrichment via HDBSCAN, Yeo-Johnson transformation, robust scaling, and class resampling using SMOTE, ADASYN, and SMOTE-ENN. Four ensemble classification models (Random Forest, XGBoost, AdaBoost, and LightGBM) were evaluated. LightGBM with SMOTE yielded the highest recall (0.99) and the fewest false negatives. These findings suggest a solution for a proactive early warning system risk mitigation in electricity under extreme weather conditions.

Design and Development of an Electric Power Mirror Simulator on Vehicles

2026-02-03T11:37:05+07:00

The electric power mirror is one of the essential components in modern vehicles. Understanding how the electric power mirror works is crucial, as it facilitates troubleshooting and repair when the system malfunctions. The development of an electric power mirror simulator aims to help users directly understand the components, working principles, and electrical circuits involved, through an interactive simulator board. This simulator was created using an experimental method involving the design, frame construction, material collection, and testing processes. The test results show that the electric power mirror simulator functions properly and is suitable for use as an educational tool, The average voltage flowing in the circuit is 11.4 volts, and the electric current is 1.48 amperes.

Implementation of the RAD Method in a Point of Sales Information System with Payment Gateway Integration at Kedai Ngopi Ngopi

2026-02-03T11:37:05+07:00

Kedai Ngopi Ngopi faces several operational challenges, including the absence of a digital menu display, manual transaction recording, and limited payment methods. This study aims to develop a web-based information system integrated with sales and digital payment functionalities. The system development was carried out using the Rapid Application Development (RAD) method, which emphasizes speed and active user involvement through iterative processes. The developed system includes product management, transaction recording, order notifications, and integration with a digital payment system. Testing results show that the system significantly improves operational efficiency, expands available payment methods, and enhances the overall customer service experience.

Optimization of Perturb and Observe Algorithm Efficiency in MPPT PV Array System Using Boost Converter: A Simulink-Based Investigation

2026-02-03T11:37:05+07:00

The increasing demand for renewable energy requires efficient power conversion and control strategies to maximize the utilization of photovoltaic (PV) systems. Among various approaches, Maximum Power Point Tracking (MPPT) is critical in ensuring optimal energy extraction under varying irradiance and load conditions. This study focuses on optimizing the efficiency of the Perturb and Observe (P&O) algorithm implemented in a PV array system with a DC–DC boost converter. A comprehensive Simulink-based investigation is conducted to analyze the algorithm’s performance in terms of tracking speed, stability, and steady-state oscillation. The results demonstrate that the optimized P&O method significantly improves the dynamic response of the boost converter while maintaining high tracking accuracy under irradiance fluctuations and load variations. Furthermore, this work is presented as a preliminary study prior to the integration of MPPT control with a PLC-based monitoring system, providing a reliable foundation for real-time industrial applications. The findings contribute to the enhancement of PV system efficiency and offer valuable insights for future implementation in smart energy management platforms.

Digital Forensics Implementation on CCTV Using ISO/IEC 27037 and ISO/IEC 27042

2026-02-03T11:37:05+07:00

Digital evidence in the form of CCTV footage is crucial data in legal investigations, but its integrity and authenticity are often questioned due to the lack of standardization of digital forensic procedures. This study aims to design and test the effectiveness of the implementation of a digital forensic framework based on the international standards ISO/IEC 27037:2012 (guidelines for acquisition and preservation) and ISO/IEC 27042 (guidelines for analysis and interpretation). The research method involves the application of the four-stage procedural ISO/IEC 27037 (Identification, Collection, Acquisition, and Preservation) on a CCTV V380 MicroSD card, followed by analysis using ISO/IEC 27042 principles and Autopsy software. The results show that this framework is fully effective. The integrity of the digital evidence is guaranteed, evidenced by the consistency of hash values (MD5 and SHA1) that are verified Matched throughout the acquisition process until the preparation for analysis. Furthermore, the analysis procedure successfully retrieves information from videos that have been deleted from Unallocated Space, complete with important forensic metadata. This study concludes that the ISO/IEC framework has proven effective in ensuring the authenticity of evidence, and is therefore recommended as a scientifically and legally valid standard operating procedure model for forensic practitioners.

Web Application Security Testing Against SQL Injection Attacks Using SQL Map

2026-02-03T11:37:04+07:00

Web application security is a major concern due to the increasing threat of cyberattacks, especially SQL injection attacks, which threaten the integrity, confidentiality, and availability of data. This study aims to measure the vulnerability of web applications to SQL injection attacks using penetration testing methods. This test is carried out using a tool in the form of SQLMAP that can detect and exploit vulnerabilities through the boolean-based blind SQL injection technique and the error-based injection technique. In addition, this study also implements and tests the protection ability using the addlashes()- based input filtering method in PHP. The test results show that SQLMAP is widely used because protection against this vulnerability can provide a robust solution on how to protect web applications. That way, web applications are expected to be safe from attacks that damage existing data and systems

Honda Vario 125 Engine Modification Performance Analysis

2026-02-03T11:37:04+07:00

This study aims to describe the performance of the Honda Vario 125 engine that has undergone modifications to five main components, namely the piston, valve, camshaft, throttle body, and injector. Performance measurements were carried out to determine changes in torque and horsepower after the modifications were applied. The research method used was a quantitative descriptive approach with testing using a dyno test. The test procedure included engine preparation, tool calibration, and the implementation of a full dyno test from idle to maximum speed. The results showed that the engine was capable of producing a maximum torque of 31.36 N m at 3,507 rpm and a maximum power of 20.21 HP at 7,693 rpm. These values indicate an increase in the engine's ability to produce power and thrust at certain speeds. Overall, this study provides an empirical description of the performance characteristics of the Honda Vario 125 engine after mechanical modifications

Analytical Hierarchy Process (AHP) for Landslide Vulnerability Zoning in the Gumelar Area

2026-02-03T11:37:04+07:00

Landslides are geological disasters that frequently occur in areas with steep topography and minimal vegetation, including in Gumelar Subdistrict, Banyumas Regency. This study aims to map landslide vulnerability zones using a multi-criteria approach with the Analytical Hierarchy Process (AHP) method integrated into a Geographic Information System (GIS). Four main parameters analyzed include slope gradient, rainfall, lithology, and land cover, with weights determined through a pairwise comparison matrix by experts. The results indicate that slope gradient (49.2%) and rainfall (30.9%) are the dominant factors in determining vulnerability levels. The resulting vulnerability map shows the distribution of areas with low, moderate, and high risks, validated using field landslide event data. This study provides an accurate spatial basis for landslide disaster mitigation planning in the study area

Implementation of Agile Modeling In Web-Based Quality Management Audit Application

2026-02-14T11:38:14+07:00

Information Technology Certification Center (ITCC) is an institution responsible for implementing certification at ITPLN, with a focus on certification in the field of technology. Currently, ITCC still uses a conventional system in managing its archives. To overcome this, a web-based quality management audit application has been designed using the Agile approach. The Agile approach was chosen because of its ability to increase development efficiency, provide flexibility in customizing applications according to needs, and accelerate project completion. The resulting application is also more adaptive to user feedback and supports a dynamic audit process. The development of this application involves the use of the PHP programming language, MySQL database, and testing using the Blackbox testing method.

Implementation of Augmented Reality (AR) as a Learning Medium for Introducing Hijaiyah Letters (A Case Study at UMP Kindergarten)

2026-02-03T11:37:04+07:00

In today's digital era, Augmented Reality (AR) technology is increasingly being utilized as a learning medium to increase children's interest in learning, including in introducing hijaiyah letters to kindergarten children. The problem that is often faced is the lack of interactive educational applications that are suitable for this age. This research aims to develop AR based educational applications using the Multimedia Development Life Cycle (MDLC) method, which includes six stages, namely concept, design, material collection, manufacture, testing, and distribution. The application was tested using the black box testing method on various Android devices. The test results show that all features such as camera access, marker scanning, 3D appearance of hijaiyah letters, interactive quizzes, instructions for use, and author information function properly. This application supports learning media that is interactive, fun, and helps children recognize hijaiyah letters better.

Implementation of Midtrans Payment Gateway on the Zumii Store Game Top-Up Platform Using the Laravel Framework

2026-02-03T11:37:04+07:00

This study discusses the development of a responsive website for the Zumii Store Game Top-Up Platform using the prototyping method. The Research was motivated by the need to improve the efficiency of top-up transaction, which were previously handled manually and were prone to eror. The system was developed iteratively using the prototyping approach involving direct user feedback at every stage. The result show that the implemented system can facilitate faster, more accurate, and easily accessible digital ordering, order tracking, and payment processes via both desktop and smartphone devices. Integration with the Midtrans payment gateway also accelerates payment verification and automates product delivery. The adoption of responsive web design and prototyping method has been proven to improve quality and minimize transaction error on the Zumii Store game Top-Up platform.

Application of the Constant Current Method in Lead-Acid Battery Charging

2026-02-03T11:37:05+07:00

Air pollution caused by fossil fuel-powered vehicles has a significant impact on public health and the environment, encouraging the transition to electric vehicles in accordance with Presidential Regulation No. 55 of 2019. Electric vehicles use batteries as their main energy source, making a safe charging system essential to prevent overcharging. This study designs a lead-acid battery fast charging system based on the Constant Current (CC) method with a constant current of 2.7 A, equipped with an automatic cut-off. Testing was conducted through field observation, interviews, and literature review. Results show that the charging time is only ±3 hours 13 minutes, significantly faster than the conventional method (10 hours), with component measurements within tolerance limits. The PZEM-017 and voltage divider sensors accurately read voltage according to the datasheet. This system proves to be efficient, safe, and capable of preventing battery damage due to overcharging.

Warehouse Re-Layout for a Third-Party Logistics Provider Based on Shipping Patterns Using the Apriori Algorithm

2026-02-03T11:37:04+07:00

Warehouse operations face challenges that affect efficiency. PT ABC, a third-party logistics provider, has been experiencing these issues due to an unstructured warehouse layout. It leads to inefficient movement and delayed deliveries. This study aims to re-design the warehouse layout of PT ABC based on shipping patterns using the apriori algorithm. Shipment data from January to June 2024 was analyzed using the apriori algorithm with a minimum support of 30% and a confidence level of 50% to find frequently shipped item combinations. One of the strong associations found was between "Mechanical Components" and "Spare Parts," with a lift value of 1.68. Based on the results, a proposed layout was developed by considering both the FSN (Fast, Slow, Non-moving) classification and apriori findings. The effectiveness of the new layout was evaluated using rectilinear distance calculations, which showed a reduction in total item movement distance from 45,952 meters to 44,343 meters.

Optimization of Photovoltaic Performance Through the Implementation of a Photodiode-Controlled Dual-Axis Solar Tracking Mechanism

2026-02-03T11:37:04+07:00

Indonesia has abundant renewable energy resources that support efforts toward energy independence and long-term energy security. However, photovoltaic (PV) modules are commonly installed in a fixed orientation, which limits solar energy absorption as irradiance varies with the sun’s position. To address this issue, this study employs a photodiode-based dual-axis solar tracking system capable of continuously aligning the panel with the sun’s trajectory. The photodiodes detect the direction of incoming sunlight, enabling real-time orientation adjustments to improve energy harvesting. Experimental results show that the dual-axis tracker produces an average output of 20.51 V and 0.98 A, compared to the static PV module with 19.97 V and 1.07 A.

Design and Implementation of an IoT-Enabled Deep Learning Vision System for Automated Dimensional Measurement in Smart Manufacturing

2026-02-03T11:37:04+07:00

The rapid advancement of Industry 4.0 has brought the convergence of Internet of Things (IoT), computer vision, and deep learning to enhance automation and precision in manufacturing. This paper presents the design and implementation of an IoT-enabled deep learning vision system for automated dimensional measurement, integrated with programmable logic controller (PLC) control and real-time monitoring. The system employs a Raspberry Pi 5 as an edge computing unit, Logitech C270 camera for visual data acquisition, and an Omron CP2E PLC for process control. A YOLOv5 deep learning model is trained to detect and measure object dimensions with sub-millimeter accuracy. The Node-RED platform is utilized for dashboard visualization and communication, interfaced through Omron FINS protocol, with MySQL as the database for data logging. Experimental results show a high detection accuracy of 98.6% and an average measurement error of less than 0.5 mm, demonstrating the system’s effectiveness for smart manufacturing applications.

Comparative Analysis of Expert System Methods for Early Diagnosis of Online Game Addiction A Systematic Review

2026-02-03T11:37:04+07:00

The phenomenon of online game addiction has become a serious problem that requires early detection and prevention. This study aims to analyze and compare several expert system methods that have been applied to diagnose online game addiction, to provide an overview of the most effective and efficient methods. The methods analyzed include Backward Chaining, Forward Chaining, Certainty Factor, Euclidean Distance, Fuzzy Tsukamoto, Fuzzy Sugeno, Fuzzy Mamdani, Case-Based Reasoning, and K-Nearest Neighbor. Based on a review of 15 literature sources, the results of the analysis show that the diagnostic accuracy of each method varies. Certainty Factor achieves an accuracy of 79-81.2%, the combination of Certainty Factor and Forward Chaining reaches 99.64%, Backward Chaining and Certainty Factor at 80%, while Fuzzy Tsukamoto with Fuzzy Sugeno and Case-Based Reasoning is able to achieve perfect accuracy of 100%. The conclusion of this study shows that hybrid methods that combine fuzzy inference and reasoning techniques, such as Fuzzy Tsukamoto-Fuzzy Sugeno and Case-Based Reasoning, show a very high level of accuracy and have the potential to be the most effective approach for an expert system for diagnosing online game addiction, mimicking the diagnostic capabilities of a psychologist.

Utilization of LLM in the Automation Process of Contract Template Recognition

2026-02-03T11:37:05+07:00

This research investigates the use of various text similarity methods in automating the recognition of varied contract templates. Determining the correct template is a crucial step before the automation process proceeds to the clause-by-clause evaluation stage. This recognition process involves dynamically comparing clause text between drafts and templates without data labeling, relying on available text. Testing was conducted using traditional methods (Jaccard similarity, TF-IDF, BM25) and natural language processing methods (BERT, LaBSE, LLM). The research methodology involves acquiring contract samples from various sources, creating templates, and testing template recognition. The testing output is evaluated based on its effectiveness in capturing semantic equivalence and contextual understanding. Research results show that LLM is highly robust in recognizing templates by only learning from the first few sample clauses. These findings indicate that template recognition automation through LLM will provide the best precision and accuracy compared to traditional methods and other natural language processing methods. Thus, this research can serve as a foundation for developing a template-based contract review automation system that is more robust against contract variations.

Development of Digital Twin Modeling for Smart Factory using OpenPLC-Based Control System

2026-02-03T11:37:05+07:00

This research discusses how to design and develop Digital Twin modeling in a Smart Factory system with an OpenPLC-based control system. The main problems in conventional manufacturing systems are limited production line flexibility, difficulty in monitoring and controlling complex devices, and process visualization that is limited to technical data without a real depiction. Through the design of a laboratory-scale mechanical system, the creation of a control board based on an STM32H7 microcontroller with an OpenPLC platform, and the development of a Digital Twin application using Three.js and Node.js, this research produces a Smart Factory prototype that can be visualized in three dimensions and operated in real time via a local network or the internet. Test results show that the system is able to display sensor conditions, control actuators, and accelerate decision-making with shorter response times. The implementation of this system supports production process efficiency and offers an alternative industrial automation solution that is economical, flexible, and more intuitive in the implementation of digital transformation.

Mechanical Hazard Simulator of PLC-Based Automation Technology According to ISO 12100

2026-02-03T11:37:05+07:00

Automation technology is expected can enhance productivity, reduce unnecessary cost, and avoid the potential upcoming accidents. However, many automation machineries don’t fulfil safety requirements and have no third party safety certification. This situation requires internal risk assessment as a gate before technology deployment in order to reduce potential hazardous events. This research applies internal risk assessment of Programmable Logic Controller (PLC)-based automation technology according to ISO 12100. Mechanical hazard simulator is built to describe the potential hazard and estimate risk level. Alternative solutions for technology retrofit is selected and built to reduce risk level. At the end of the research, risk level comparison is presented as the evidence of risk level reduction.

Comparative Analysis of ABS Braking Performance Using Fuzzy Logic Controller and Bang-Bang Control under Dry and Wet Road Conditions

2026-02-03T11:37:04+07:00

In this study, The effectiveness of the Anti-Lock Braking System is evaluated by comparing bang-bang control with fuzzy logic control. ABS plays a crucial role in vehicle safety, as it manages braking torque to ensure wheel slip remains at an optimal level, preventing the wheels from locking and maintaining control of the vehicle. A mathematical ABS model was implemented and analyzed using MATLAB/Simulink simulations to examine how both controllers behave. The simulation outcomes highlight that fuzzy logic control performs better than bang-bang control, with advantages such as shorter stopping distances, faster braking responses, and more stable slip control.

Design and Development of a Microcontroller-Based Catfish Fry Counter

2026-02-03T11:37:05+07:00

Catfish farming in Indonesia has significant potential; however, conventional methods of counting fish fry still face various challenges, such as the considerable time required and a high risk of errors. This study aims to design and build an automatic counting device for catfish fry based on the Arduino Mega 2560 microcontroller, integrated with an E18-D80NK infrared proximity sensor and a DS18B20 temperature sensor. The device is specifically designed to count catfish fry measuring between 6 to 8 cm, where the infrared proximity sensor is used to count the number of fish fry passing through, while the DS18B20 sensor functions to detect the water temperature in the holding tank. The operational process of the device includes inputting the number of fry via a 4x4 keypad, opening the counting path using a servo motor, measuring the water temperature, displaying data on a 16x2 LCD, and closing the path after the counting process is completed. The test results indicate that the DS18B20 sensor has an accuracy of 97.68% with an average error of 2.32%, while the infrared proximity sensor achieves an accuracy of 89.08% with an average error of 10.92%. This device successfully counts the number of catfish fry automatically, monitors the water temperature, and displays information in real-time on the LCD. Therefore, the catfish fry counting device based on the Arduino Mega 2560 is expected to enhance time and labor efficiency, as well as minimize errors in the counting and distribution processes of catfish fry.

Analysis of The Effect of Drive Pulley Diameter Variation on Electric Motorcycle Performance

2026-02-03T11:37:04+07:00

The primary objective of this study is to investigate how different sizes of drive pulleys influence the performance of an electric motorcycle equipped with a BLDC motor. This was approached using an experimental technique that combined dynotesting and road testing where three drive pulley diameters of 42T, 44T, and 46T were used. While the road tests were intended to measure the top speed of the motorcycle, the dynotests were aimed at acquiring the necessary data to calculate the torque and power. Based on the results of this study, an increase in the drive pulley's diameter will increase the speed of the electric motorcycle, although it will lower the torque. It was found that the 42T drive pulley offers the greatest torque, which is 16.54 Nm at a power of 5.2 HP, and the 46T drive pulley allows for the greatest speed of 63.37 km/h. It is thus apparent that for performance to be properly tailored to the motorcycle, a balance should be struck in line with the intended purpose, which can either be acceleration or the cruising speed.

Study of the Effect of Simple Filtration Composition of Waste Cooking Oil on the Quality of Biodiesel for Generator Fuel

2026-02-03T11:37:04+07:00

Used cooking oil has great potential as a raw material for biodiesel due to its abundant availability and low price, but its quality often decreases due to high free fatty acid and impurities content. This study aims to examine the effect of variations in simple filtration compositions on improving the quality of used cooking oil before the conversion process into biodiesel. The filtration process was carried out using a combination of simple media such as activated charcoal, fine sand, and zeolite in certain ratios. The parameters tested included the density value at 15°C (kg/m³), the kinematic viscosity value at 50°C (mm²/s), and the calorific value (kcal/g) of the produced biodiesel. The experimental method involved several variations in filtration compositions followed by a standard transesterification process. The results of the study showed that the optimal filtration composition was able to produce B40 biodiesel at a temperature variation of 50-60 15°C with a density value of 868.5 kg/m³, a kinematic viscosity of 3,570 mm²/s, and a calorific value of 9772.4 kcal/g according to the standard range of SNI 7182:2015. In addition, biodiesel from simply filtered used cooking oil can be used as a generator fuel with an energy efficiency performance of 34.79% in an 8.5 kVA generator test. This finding provides an alternative for simple, inexpensive, and environmentally friendly used cooking oil processing to support the availability of renewable energy.

Design and Construction of a NodeMCU-Based Heart Rate and Blood Oxygen Level Monitoring Device with Blynk Application Output

2026-02-03T11:37:04+07:00

Health is a crucial aspect of human life that must be maintained and monitored regularly, especially during recovery periods or self-isolation. Real-time and remote body monitoring has become an essential need in today’s digital era. This study aims to design and develop a heart rate and blood oxygen saturation monitoring device based on the Internet of Things (IoT) using the NodeMCU ESP8266 microcontroller integrated with the MAX30100 sensor and the Blynk application. The research method involves the design of an electronic system based on microcontrollers, software programming, and direct functional testing. The test results show that the device is capable of accurately detecting heart rate (BPM) and oxygen saturation (SpO₂), displaying the data on a 16x2 LCD, and transmitting it to the Blynk application in real-time via a WiFi connection. Therefore, the developed device has proven to function effectively as a practical and portable remote health monitoring system.

An Analysis of Students’ Perceptions and Learning Motivation in a Gamified Learning Management System

2026-02-03T11:37:04+07:00

The use of Learning Management Systems (LMS) in lectures is increasing, but student motivation to learn often poses a challenge. This study aims to examine the effect of gamification in LMS on student motivation to learn and to analyze the relationship between students' perceptions of gamification and their motivation levels. The study uses a quantitative design with measurements taken before and after the implementation of gamification, as well as correlation analysis to examine the relationship between variables. The results show that there is no significant difference between motivation before and after the implementation of gamification. However, perceptions of gamification are positively related to autonomy, connectedness, and achievement, which contribute to increased learning motivation. These findings indicate that gamification has an indirect effect through the reinforcement of psychological aspects that support intrinsic motivation.

Design And Construction of a Non-Contact Body Temperature Measuring Device Using the Mlx90614 Sensor Based on Nodemcu and the Blynk Application

2026-02-03T11:37:04+07:00

Body temperature is an important indicator for monitoring a person's health condition. The average normal human body temperature ranges from 36–37°C. Changes in body temperature, whether hyperthermia or hypothermia, may indicate infections or metabolic disorders, making accurate and practical measurement methods essential. Conventional contact thermometers have limitations, such as long measurement times and user discomfort. To overcome these issues, a non-contact body temperature measurement system was designed using an infrared sensor. This sensor works by detecting thermal radiation emitted from the body’s surface without requiring direct skin contact. The device utilizes the MLX90614 infrared sensor, an ultrasonic sensor for distance detection, and a switch as an on/off control. Measurement results are displayed on an OLED screen and can be monitored via a smartphone through the Blynk platform. Based on testing, the non-contact body temperature measurement device achieved an accuracy level of 99.3%.

Exploration of Engineered Diode Current-Voltage Convergence for Maximum Power Point Tracking in Single-Axis Solar Tracker Systems.

2026-02-03T11:37:04+07:00

Indonesia, situated in the tropics, possesses optimal solar irradiation potential, positioning it as an ideal location for the deployment of photovoltaic (PV) energy systems. Given the escalating electricity demand driven by demographic expansion and technological advancements, harnessing solar panels as a viable alternative energy source is crucial. Nevertheless, static PV installations frequently yield suboptimal power absorption. This research, titled "Exploration of Diodic Current-Voltage Convergence of Maximum Power Point Engineering in a Single-Axial Solar Tracker System," focuses on an advanced approach to power optimization. Specifically, the study designs and evaluates a system capable of optimally capturing solar radiation by integrating a 20 Watt Peak (WP) PV unit with an automatic Solaris Single-Axial Tracker. The system architecture utilizes an Arduino Nano microcontroller as the core control unit, which is equipped with an ACS712 current sensor and a DC voltage sensor. Real-time data communication is accomplished using a Node-MCU ESP8266 module for serial transmission. Comparative analysis reveals a significant performance enhancement in the PV system attributed to the automatic tracking implementation. The average power absorbed by the 20 WP solar panel increased from 89.2 W (without tracking) to 93.70 W when the Single-Axial Tracking system was employed. Correspondingly, the PV panel's efficiency improved from 78.4% (static) to 82% (tracked). Furthermore, the monitoring data (current, power, and voltage) is visualized in real-time on a smartphone via the Blynk application, enabled by an internet-based communication protocol. This investigation successfully demonstrates that a microcontroller-based single-axial tracking system offers an efficient solution for maximizing solar energy absorption in tropical environments.

Clustering for Motivating New Student Admissions in Study Program Selection: Systematic Literature Review

2026-02-03T11:37:04+07:00

This research aims to evaluate clustering in new student admissions in determining effective strategies, to help prospective students in choosing study programs that match the interests and potential of prospective new students. Clustering as a machine learning technique to group data that has similarities, is increasingly used in the field of education to support the decision-making process. This Systematic Literature Review (SLR) examines the application of clustering methods in new student admissions, especially in recommending the right study program. By analyzing 10 studies in applying clustering methods that are often used, to determine the main factors that influence the selection of courses, as well as their impact on student satisfaction in choosing courses and optimal academic results. The results of this study provide insight into strategies for the admissions team in optimizing marketing, so that there is a more effective alignment between student profiles and study program characteristics.

Design and Construction of Special Tools for Rear Seal Removal and Installation on Komatsu Engine Flywheel Housing

2026-02-03T11:37:04+07:00

This research focuses on the innovation of mechanical work aid design in the form of special tools specifically used to remove and install the rear seal on the flywheel housing of Komatsu diesel engines type SAA6D107E-1 and SAA4D95LE-3. One of the procedures that is still done manually and has high risk is the replacement of the rear seal on the flywheel housing. The purpose of using special tools is to gain efficiency in the removal and installation of the rear seal on the flywheel housing. The research method includes field observation, component measurement, 2D and 3D design, and direct field trials. The test results show that this tool can reduce the processing time from 55 minutes to 20 minutes. The use of these special tools helps technicians in increasing the work speed in removing and installing without causing damage to other components.

Modeling and Optimization of PET Filament Making Machine Control System Based on Response Surface Methodology

2026-02-03T11:37:05+07:00

This study aims to model and optimize the control system of a PET filament making machine using Response Surface Methodology (RSM). Two main variables, namely extrusion temperature and winding speed, were analyzed through the Central Composite Design experimental design with the help of Design-Expert 13 software. A quadratic model was developed to predict the filament diameter, and the results of the analysis of variance showed that the model was statistically significant and able to represent the nonlinear relationship between variables. Optimization based on the desirability function produced optimum operating conditions at a medium temperature and winding speed range. Experiments showed that the combination of a temperature of 240–245 °C and a speed of 9–12 rpm produced a consistent filament diameter of 1.7 mm with 70–80% flexibility and a smooth surface.

Causal Modeling of Factors Causing Toddler Stunting Using the Peter-Clark Algorithm

2026-02-03T11:37:05+07:00

Stunting is a growth disorder in toddlers that can result in a height that is not proportionate to their age. Between 2016 and 2024, many studies have discussed factors related to stunting. However, these studies generally only use correlation analysis, which indicates the level of closeness of the relationship between variables. Correlation analysis can indeed describe the existence of an association, but cannot explain the causal relationship. Therefore, the causal mechanisms underlying these factors have not been fully revealed. This study aims to model the causal relationship of the factors that cause stunting. This study uses the Peter-Clark algorithm to obtain the direction of the causal relationship. The results of this study show a relationship between Height(TB.U)/Weight(BB.U), Integrated Health Post Visits/Height(TB.U), Height(TB.U)/Mother's Education, Age of Marriage/Mother's Education, Immunization/Family Members Smoking in the Home, and Exclusive Breastfeeding/ Integrated Health Post Visits. The Peter-Clark algorithm in this study successfully identified a causal relationship based on a comparison of performance using directional and causal density of 67%. These results are quite informative, but 34% of the relationships remain undirected. Therefore, additional data, domain assumptions, or advanced algorithms, such as FCI or GES, are needed to determine their direction.

Support Vector Regression–Based Model for Multizone Electricity Consumption Forecasting

2026-04-04T19:22:30+07:00

This study aims to develop and evaluate a multi-zone short-term electricity consumption prediction model based on weather factors using the Support Vector Regression (SVR) method to support more efficient and adaptive power system planning in response to climate variability. Ten-minute resolution electricity consumption data from three zones was combined with variables of air temperature, relative humidity, wind speed, and solar radiation. The research process included data preprocessing, temporal feature engineering, time-based data partitioning, and SVR hyperparameter optimization with RBF kernel. Model performance was evaluated using Root Mean Square Error (RMSE), Mean Absolute Error (MAE), and coefficient of determination (R²) metrics, and compared with reference models such as linear regression, Random Forest Regression, and Artificial Neural Network. The results of the experiment show that SVR provides the best accuracy at high temporal resolution. At 10-minute aggregation, the MAPE values obtained were 4.78% (Zone 1), 4.11% (Zone 2), and 9.25% (Zone 3). Model performance declines as the level of time aggregation increases, indicating that the effectiveness of SVR is influenced by the temporal scale and load characteristics of each zone. These findings show that SVR works effectively for weather-based electricity load forecasting in various zones at high temporal resolution, although its performance declines at larger aggregation scales.

Keywords: Multizone electric power consumption; Short-term load forecasting; Weather/meteorological factors; Support Vector Regression (SVR); Intelligent energy management systems.