Publication Search

Kreo Village, located in Larangan District, Tangerang Regency, still relies heavily on very traditional communication methods in daily life. While this approach reflects local wisdom and strong cultural values, the changing times demand a digital transformation, especially in the field of communication. With the rapid advancement of digital technology and the growing quality of human resources in the village, the need for a more modern communication system is becoming increasingly important. One potential solution is the use of artificial intelligence technology such as ChatGPT. ChatGPT (Generative Pre-trained Transformer) is a language model based on AI developed by OpenAI. This technology can understand and generate natural language interactively, similar to human conversation. By integrating ChatGPT into community communication activities, Kreo Village can speed up access to information, facilitate the exchange of ideas and opinions, and bridge the existing digital gap. In addition, ChatGPT can also be used as an educational tool to help residents understand digital technology, support learning activities, and strengthen community participation in village development. Through this initiative, Kreo Village can move toward becoming a more inclusive, adaptive, and competitive digital village in the modern era.

Amyotrophic Lateral Sclerosis (ALS) is a progressive neurodegenerative disorder that presents significant diagnostic challenges due to its heterogeneous clinical manifestations and symptom overlap with other neurological conditions. Early and accurate diagnosis is critical for initiating timely interventions and improving patient outcomes. Traditional diagnostic approaches rely heavily on clinical expertise and manual interpretation of neuroimaging data, such as structural MRI, Diffusion Tensor Imaging (DTI), and functional MRI (fMRI), which are inherently time-consuming and prone to interobserver variability. Recent advances in Artificial Intelligence (AI) and Deep Learning (DL) have demonstrated potential for automating neuroimaging analysis, yet existing models often suffer from limited generalizability across modalities and datasets. To address these limitations, we propose a Transformer-augmented deep learning ensemble framework for automated ALS diagnosis using multi-modal neuroimaging data. The proposed architecture integrates Convolutional Neural Networks (CNNs), Recurrent Neural Networks (RNNs), and Vision Transformers (ViTs) to leverage the complementary strengths of spatial, temporal, and global contextual feature representations. An adaptive weighting-based fusion mechanism dynamically integrates modality-specific outputs, enhancing the robustness and reliability of the final diagnosis. Comprehensive preprocessing steps, including intensity normalization, motion correction, and modality-specific data augmentation, are employed to ensure cross-modality consistency. Evaluation using 5-fold cross-validation on a curated multi-modal ALS neuroimaging dataset demon-strates the superior performance of the proposed model, achieving a mean classification accuracy of 94.5% ± 0.7%, precision of 93.9% ± 0.8%, recall of 92.9% ± 0.9%, F1-score of 93.4% ± 0.7%, spec-ificity of 97.4% ± 0.6%, and AUC-ROC of 0.968 ± 0.004. These results significantly outperform baseline CNN models and highlight the potential of transformer-augmented ensembles in complex neurodiagnostic applications. This framework offers a promising tool for clinicians, supporting early and precise ALS detection and enabling more personalized and effective patient management strategies.

An electrostimulator is a medical device designed to deliver controlled electrical stimulation to nerves and muscles, supporting rehabilitation and therapy for patients with neuromuscular disorders. This study focuses on designing and developing a portable electrostimulator that offers three distinct waveform modes: continuous wave, discontinuous wave, and dense-disperse wave, providing versatility for different therapeutic needs. The device is powered and controlled by an Arduino Mega 2560 microcontroller, coupled with a Nextion touchscreen LCD interface that allows users to adjust waveform type, frequency, and stimulation intensity with ease. Waveforms are generated through an NE555 IC circuit, with amplitude adjusted via a potentiometer and subsequently amplified using a step-up transformer to achieve therapeutic voltage levels. Functionality and performance tests were conducted using an oscilloscope, and the device was benchmarked against a commercial KWD-808 electrostimulator. Results demonstrate that the developed electrostimulator reliably produces the intended waveforms, achieving peak voltages up to 32V and frequencies ranging from 33.3 Hz to 66.6 Hz, confirming its effectiveness and feasibility for non-clinical nerve and muscle therapy applications.

This study aims to analyze the operational condition of the distribution transformer at substation PY094, PLN ULP Pringgabaya, with a primary focus on identifying and calculating the level of load imbalance on the consumer side. Data were collected through direct measurements of electrical parameters, including voltage and current in each phase, followed by a detailed analysis of energy losses. The measurement results indicated a significant load imbalance. In Feeder B, the average phase currents were recorded at 103.8 A for phase R, 130.2 A for phase S, and 90.4 A for phase T. Meanwhile, in Feeder D, the average phase currents were 47.4 A for phase R, 18 A for phase S, and 20.4 A for phase T. This imbalance caused notable power losses in the distribution system, with an estimated daily energy loss of 28.94 kWh, assuming the system operates 12 hours per day. To address this issue, load balancing simulations were carried out using ETAP software. The simulation involved redistributing load values across each phase in the two main feeders. Feeder B was simulated at 46.82% of the transformer’s full capacity, while Feeder D was simulated at 12.38% of the total 160 kVA capacity. The simulation results demonstrated that redistributing the load significantly reduced the current imbalance, thereby minimizing power losses and improving the operational efficiency of the distribution substation. Therefore, load balancing strategies are essential for enhancing energy efficiency and ensuring the reliability of electricity supply in distribution networks.

This study aims to analyze the effect of transformer oil purification on oil breakdown voltage and evaluate its benefits in maintaining transformer performance and reliability. Oil purification is a crucial preventive maintenance step to preserve insulation quality and prevent operational failures caused by reduced dielectric properties. The study was conducted through several stages, including data collection, oil purification, measurement of breakdown voltage before and after purification, and evaluation of results. Data collection involved measuring the oil’s electrical properties according to SPLN 49-1:1982 and observing results using a Break Down Voltage (BDV) test. Purification of a Trafindo 400 kVA transformer was performed through visual inspection, connecting the inlet and outlet hoses to the purification machine, and circulating the oil until the breakdown voltage met the required standards. Results indicated that the oil breakdown voltage before purification was below standard due to reduced insulation quality caused by water contamination, charcoal particles, thermal degradation, and dissolved gases that weakened dielectric properties. Additional factors such as electrical stress, mechanical stress, and excessive loading also contributed to insulation deterioration. After purification, the oil breakdown voltage increased significantly to meet the standard of >30 kV/2.5 mm, demonstrating that purification effectively restores the oil’s insulating capacity and supports optimal transformer performance. Analysis confirms that the transformer oil remains suitable for use, and routine annual purification is recommended to maintain reliability, efficiency, and operational performance. This study highlights oil purification as an effective preventive measure for transformer stability, extending operational life, and reducing the risk of insulation failure. The findings provide valuable guidance for transformer maintenance in the electricity industry, ensuring safe and optimal long-term operation.

Chat Generative Pretrained Transformer (ChatGPT) has rapidly gained popularity and is increasingly utilized across various fields, including education, where it plays a significant role as a supporting tool for academic tasks. In the educational context, ChatGPT can assist students in preparing and completing a wide range of assignments, providing quick access to structured information and alternative perspectives. This study was designed to explore the views of both students and lecturers on the integration of ChatGPT in academic activities, specifically as a support system for handling student assignments. Employing a descriptive qualitative approach, data were collected through interviews with 5 students and 6 lecturers from the Business Education Study Program at the State University of Surabaya. The findings reveal contrasting yet complementary perspectives. From the students’ point of view, ChatGPT offers considerable benefits, particularly in enhancing independence, efficiency, and time management when working on lecture assignments. However, these advantages are accompanied by drawbacks, such as decreased creativity, reduced critical thinking, and diminished interest in consulting traditional learning resources like books. On the other hand, lecturers acknowledge the usefulness of ChatGPT in simplifying student work and accelerating the completion of assignments but express concerns regarding the overreliance on artificial intelligence, which could potentially hinder the development of essential academic skills. Overall, the research suggests that while ChatGPT presents valuable opportunities to enhance learning processes, it must be applied thoughtfully, with careful guidance from educators to balance efficiency with the cultivation of creativity, critical thinking, and academic integrity.

Understanding electric motors and transformers is crucial in the field of electrical power engineering education. To enhance practical learning, a trainer was designed for single-phase AC motors, incorporating an Earth Leakage Circuit Breaker (ELCB) safety system to prevent electric current leakage, which could pose a risk during experiments. The aim of this research is to design and develop a trainer module that serves as both an interactive and safe learning tool for laboratory practicums. The trainer is equipped with a safety circuit, along with controllers such as a voltmeter, ammeter, selector switch, pushbutton, relay, and potentiometer, allowing students to understand and control various electrical parameters. The tool underwent voltage measurements and functional tests at three distinct measurement points to evaluate its performance and safety features. The results demonstrated that the trainer performed effectively, significantly enhancing students' practical understanding of electric power systems and contributing to better hands-on learning experiences in electrical engineering.

Information technology refers to any technology that assists human needs. Advances in computing and information technology significantly impact education through the use of computer networks and the internet. Artificial Intelligence (AI) is one of the most beneficial applications across various fields, including healthcare, industry, governance, and particularly in education, as exemplified by ChatGPT (Chat Generative Pre-Trained Transformer). This study focuses on the accuracy of ChatGPT in analyzing learning materials for Buddhist Religious Education. It not only examines the technological accuracy but also explores its potential as an effective learning tool in formal educational settings. The method employed is a literature review, aiming to investigate the accuracy of ChatGPT in educational contexts. This research seeks to evaluate ChatGPT's capability to analyze materials and provide precise answers. However, as ChatGPT continues to evolve, there is a risk of increasing student dependency on its use. Despite its advantages, ChatGPT also has limitations, such as its reliance on available data.  

The Sundanese language, once spoken by 48 million individuals, has experienced a significant decline in speakers, losing 2 million in the past decade. This decline is attributed to weakened intergenerational transmission and the dominance of more widely used languages. The challenges in developing Natural Language Processing (NLP) tools for Sundanese stem from the lack of annotated corpora, trained language models, and adequate processing tools, complicating efforts to preserve and enhance the language's usability. This research aims to address these challenges by implementing emotion classification in Sundanese text using Long Short-Term Memory (LSTM) and Bidirectional Encoder Representations from Transformers (BERT) models. The study utilizes a dataset of annotated Sundanese tweets, applying preprocessing techniques such as cleansing, stopword removal, stemming, and tokenization to prepare the data for analysis. The results indicate that the BERT model significantly outperforms the LSTM model, achieving an accuracy of approximately 80% compared to LSTM's 70%. These findings highlight the potential of advanced NLP techniques in enhancing the understanding of emotional nuances in Sundanese communication and contribute to the revitalization of the language in the digital age.

This study aims to analyze the sentiment of iPhone product reviews fromAmazon using the BERT (Bidirectional Encoder Representations from Transformers) model to classify reviews as either positive or negative. The dataset, sourced from Kaggle, includes text reviews and star ratings, where high ratings indicate positive sentiment and low ratings indicate negative sentiment. After text preprocessing steps, including data cleaning, tokenization, and sentiment labeling, the BERT model was fine-tuned for sentiment classification, with the data split into training, validation, and test sets. Evaluation results demonstrate that the BERT model achieves a high classification accuracy, with an accuracy rate of 93.9% and a balanced F1 score between precision and recall. Confusion matrix evaluation also indicates that the model consistently identifies both positive and negative sentiments. This study shows that Transformer-based models like BERT are highly effective in understanding customer opinions in e-commerce, with broad application potential for data-driven decision-making in marketing strategies and product development.

PT Mass Rapid Transit Jakarta operates a mass transportation system from Lebak Bulus Station to Bundaran HI. One of the traction substations is located in Cipete Raya, with a voltage rating of 20kV/1.2kV. A critical piece of equipment in this substation is the traction transformer, with a capacity of 4850 kVA/2x2500 kVA. The purpose of this study is to predict the service life of the Cipete Raya traction transformer based on temperature and load using the linear regression method. This study employs direct observation, analyzing load data from traction transformers 1 and 2 at Cipete Raya from January 2022 to June 2024, along with transformer temperature measurements. Secondary data include the technical specifications of the Cipete Raya traction transformer. The linear regression analysis for transformer 1 yields the equation y = 687.42 + 11.97x, indicating a 5.75% annual increase over the next 5 years, with a very strong correlation coefficient of R = 0.919. For transformer 2, the equation is y = 815.4543 + 6.488x, showing a 3% annual increase, with a strong correlation coefficient of R = 0.814. Based on the transformer aging calculations for June 2024, Transformer 1 has a per unit aging value (V) of 0.0014 and an estimated service life (n) of 407.689 years, while Transformer 2 has a V of 0.0012 and an estimated service life of 496.77 years. The aging model evaluation using MAPE shows that the prediction accuracy for transformers 1 and 2 is 6% and 3%, respectively, indicating excellent modeling performance.    

Energy Electricity is something that is needed in people's lives, the need for electrical energy is increasing. the need for electrical energy is increasing. demand for electrical energy continues to increase along with economic growth and community welfare. the welfare of society. The growth in demand for electrical energy is influenced by the development of the growing manufacturing and industrial sectors. This study aims to determine the energy imbalance between the industrial and general power lines on the reliability of the 20 kV network. Methods research methods used in this study are primary and secondary research. secondary research. The results obtained in this study are the imbalance of load on the 20 kV power load on the 20 kV industrial repeater transformer power at the time of high loading during the day is 192.4633333 A with a percentage of the power of the 20 kV network. day 192.4633333 A with a percentage of 0.556252408 % and the lowest general repeater load of 68.45 A with a load percentage of 0. %percentage of 0.197919075%.    

This study aims to find out about the procedure for measuring the isolation resistance of the main equipment of the 150 kV conventional substation (AIS) as many as 5 bays at the Panakkukang Substation through the measurement of the insulation resistance value in accordance with the predetermined standards. The method used in this study is quantitative descriptive. The subject of the study to be studied is a comparison of isolation resistance values in pre- and post-treatment. This was done to find out if there was a change in the insulation resistance value of the main equipment of the Substation. The main equipment studied is LA, PMT, CT, VT, and Transformer. Measurements were taken with a Megger measuring instrument, with a voltage injection of 5kV. For 5kV voltage injection, the maximum value of the test device is >1000 GΩ. The standard used is IEEE 43-2000, which is > 1 MΩ / 1 kV. Based on the results of research that has been carried out on each main equipment of the substation at GI Panakkukang. It can be concluded that the isolation resistance value in each main equipment of the substation at GI Panakkukang is above the minimum value (> MΩ/1 KV) in accordance with the IEEE Std 62: 1995 standard, VDE Catalogue 228/4, and the standard used by PLN in the "Power System Maintenance Guidebook" in 2014 the isolation resistance value at each main substation at GI Panakkukang has increased after being maintained. Maintenance can be carried out by cleaning the equipment body and foreign materials, and measuring the insulation resistance periodically.

Growing demand for electricity savings has led to the development of an automatic LED emergency light system. It is based on providing light when the power is cut off. Once fully charged, the battery ceases charging, and in the event of a power failure, the LEDs are automatically powered by the battery. The project focuses on two primary functions: it automatically activates during power outages to give illumination, eliminating the need to search for the switch, and the battery rapidly begins recharging as the main power is restored. The emergency light is crucial due to the inconsistent voltage distribution and frequent power outages in operational regions of communities and diverse enterprises. The system includes a power supply that converts 230V AC to 12V DC, a relay that uses a control pulse to alternate between connecting the battery to the LEDs and isolating it, and a rechargeable Li-ion battery that supplies power to the LEDs during blackouts. The parallel-connected LEDs light up during a power outage in the circuits. The circuit architecture shown here serves to mitigate the entire discharge of the battery, hence enhancing the battery's longevity. Key components of the system include a step-down transformer, a bridge circuit to convert AC to DC, a Zener diode to maintain voltage stability, capacitors for energy storage, and various diodes to control current flow. The project highlights the advantages of LED emergency lights, such as efficiency, longevity, and minimal energy waste, though it acknowledges the higher initial cost and temperature sensitivity as disadvantages. The automatic LED emergency light is suitable for use in homes, offices, retail shops, and other commercial settings. The project demonstrates a cost-effective and compact solution that enhances daily life by providing reliable lighting during power failures.

The advancement of e-commerce has changed the way people shop. However, there is a mismatch between the actual quality of a product and the seller’s description. Product reviews are an important source of information for making purchasing decisions. However, processing large numbers of reviews manually is difficult. This research aims to detect emotions in Indonesian language product review texts using contextual embeddings. The public dataset used was PRDECT-ID, which comprises five emotion labels. The methods used include data preprocessing, feature extraction using contextual embeddings such as Bidirectional Encoder Representations from Transformers (BERT), and classification using Decision Tree, Naïve Bayes, and k-Nearest Neighbors (KNN). Among the compared models, the KNN model demonstrated the highest improvement, achieving a 15.09% enhancement over the decision tree results. This research provides insights into the effectiveness of contextual embeddings in detecting emotions in Indonesian language product review texts.

Generative Pre-trained Transformer (ChatGPT), is an artificial intelligence (AI) model that has revolutionized the educational landscape in recent years. Its use has been explored across various educational fields; however, given that this AI model is language-based, it has attracted particular interest in language teaching, especially in developing the writing skill. Due to the importance of the writing skill in language learning, and the extensive body of research investigating the best ways to support this skill using ChatGPT, the current study aims to provide an overview of the latest findings regarding the use of ChatGPT in developing writing. Through a narrative literature review approach, this paper will offer comprehensive documentation of the various ways ChatGPT can be used to support the writing skill, as well as the potential risks associated with its implementation. These objectives will be achieved by addressing two key research questions: In what ways can ChatGPT be useful in developing the writing skill? and what are the possible risks of implementing this tool in developing the writing skill? By answering these questions, the paper will help equip educators and learners with a comprehensive understanding of the most effective ways to use this tool to enhance the writing skill.

Obtaining high-quality, diverse, accurate datasets for sentiment analysis has always been a significant challenge. Traditional approaches include annotators, which may introduce bias to datasets and are also time-consuming and expensive. These types of datasets may also not represent the variety needed to train robust and generalizable sentiment analysis models. This study introduces a novel combination of techniques to approach the problem with a novel solution. The proposed system, SentiGEN includes the use of a transformer, T5, fine-tuned and optimized using an evolutionary algorithm to generate high-quality, diverse, accurate data for sentiment analysis. The generated data is validated using XLNet to ensure high sentiment accuracy. This combination of technologies has proven successful based on the results derived from evaluating multiple models. From complex transformers such as BERT to more straightforward approaches like KNN, those trained using synthetic data demonstrated superior performance compared to their counterparts trained on real data. This enhancement in predictive accuracy was observed when evaluated on benchmark datasets such as SST-2 and Yelp. SentiGEN can generate high-quality, diverse, accurate, realistic data for sentiment analysis and successfully increased the performance of models trained on synthetic data compared to the same model trained on real data.

In the evolving landscape of agricultural technology, recognizing rice diseases through computational models is a critical challenge, predominantly addressed through Convolutional Neural Networks (CNN). However, the localized feature extraction of CNNs often falls short in complex scenarios, necessitating a shift towards models capable of global contextual understanding. Enter the Vision Transformer (ViT), a paradigm-shifting deep learning model that leverages a self-attention mechanism to transcend the limitations of CNNs by capturing image features in a comprehensive global context. This research embarks on an ambitious journey to refine and adapt the ViT Base(B) transfer learning model for the nuanced task of rice disease recognition. Through meticulous reconfiguration, layer augmentation, and hyperparameter tuning, the study tests the model's prowess across both balanced and imbalanced datasets, revealing its remarkable ability to outperform traditional CNN models, including VGG, MobileNet, and EfficientNet. The proposed ViT model not only achieved superior recall (0.9792), precision (0.9815), specificity (0.9938), f1-score (0.9791), and accuracy (0.9792) on challenging datasets but also established a new benchmark in rice disease recognition, underscoring its potential as a transformative tool in the agricultural domain. This work not only showcases the ViT model's superior performance and stability across diverse tasks and datasets but also illuminates its potential to revolutionize rice disease recognition, setting the stage for future explorations in agricultural AI applications.

Power systems are critical infrastructure that face significant challenges due to increasing demand and inherent complexity. Predicting failures in power systems is crucial for enhancing grid reliability, minimizing downtime, and optimizing maintenance processes. This study evaluates various deep learning models, specifically convolutional neural networks (CNN), recurrent neural networks (RNN), and transformer models, for predicting power system failures. By analyzing these models’ performance metrics on historical power grid data, the study provides insights into the strengths and weaknesses of each approach. The findings contribute to the development of more robust predictive models for power system reliability.

Distribution transformers have a very important role in the electric power system. The lifespan of transformers in electric power systems will decrease over time. The reduced service life of distribution transformers is caused by several factors, including loading, ambient temperature, transformer winding temperature and transformer oil temperature. The author uses quantitative and qualitative research types. This research was conducted to determine the remaining useful life of distribution transformers. The data used in this research are transformer template data and peak load data during the day and night. The results of data processing obtained state that the first transformer's estimated remaining life with a load > 80% is 18 years starting from 2023, the second transformer's estimated remaining life with a load > 80% is 16 years starting from 2023, for the third and fourth transformers it is not calculated because it has been used since 1982 or is around 41 years old. For the winding temperature on the first transformer LBP 84 ℃ and BP 89 ℃, on the second transformer the winding temperature LBP 81 ℃ and BP 96 ℃. Both transformers are still considered good because the hot spot temperature is below the maximum limit set by the IEEE in 1955, namely a temperature of 98 ℃.