Jurnal Politeknik Negeri Batam (PoliBatam)
Not a member yet
3001 research outputs found
Sort by
Comparing Decision Tree and Support Vector Machines in Hospital Satisfaction
Get PDFPatient satisfaction is a key indicator of hospital service quality. This study compares the performance of Decision Tree and Support Vector Machine (SVM) in classifying patient satisfaction at Harapan Hospital Magelang for service optimization. The dataset, derived from a 2024 survey, consists of 577 samples and 13 predictor variables, covering patient demographics and medical service aspects. Preprocessing includes data cleaning, normalization, encoding, and class balancing using SMOTE. The Decision Tree is applied with gini impurity and max_depth=11, while SVM uses the RBF kernel (C=100, gamma=0.01). Model evaluation metrics include accuracy, precision, recall, F1-score, and ROC-AUC.Results show that Decision Tree outperforms SVM, achieving 86% accuracy vs. 81%. It also has 86% precision and 95% recall for the Dissatisfied category, higher than SVM (93% recall). The McNemar test confirms a statistically significant performance difference (p-value = 0.037). With higher accuracy and interpretability, Decision Tree is recommended as the primary method for hospital patient satisfaction analysis. These findings support the development of an adaptive classification system for Indonesian healthcare data
System Usability Scale (SUS) Model in Evaluating Internal Quality Audit Systems for Accreditation Process Optimization
Get PDFThe usability of internal quality audit systems plays a crucial role in ensuring the effectiveness of academic governance in higher education institutions. This study evaluates the usability of an internal quality audit system using the System Usability Scale (SUS) framework. The findings reveal a mean SUS score of 72.4, indicating good usability but highlighting key areas for improvement, particularly in navigation, interoperability, and data management workflows. A comparative analysis with previous research shows that while the system performs better than some older academic management systems, it still lags behind general academic information systems. Based on user feedback, recommendations for interface redesign, system integration, and workflow optimization are proposed to enhance overall usability. Future research should focus on user-centered design improvements, AI-driven automation, and broader system interoperability to further refine the effectiveness of quality audit processes in higher education institutions
The Impact of Work Stress and Job Burnout on Turnover Intention among Indonesia-China Integrated Industrial Employees
Get PDFThis research aims to analyze the impact of work stress and job burnout on Turnover Intention in companies located in the Indonesia-China Integrated Industrial Zone. A quantitative research method is used. Purposive sampling was a sampling technique involving 107 respondents. Data sources come from primary and secondary data. Primary data is distributed in the form of a questionnaire via Google Forms as an intermediary. Meanwhile, secondary data comes from internal case study companies. Closed statements were submitted in this study, and a Likert scale was used to indicate strongly disagree, disagree, agree, and strongly agree. Partial Least Square is a technique for data analysis using Smart PLS version 4.1.0.0 as a data processing tool. the first hypothesis indicates a relationship between work stress and turnover intention. Turnover intention and job burnout are unaffected by each other, as the second hypothesis demonstrates. Workers need to be able to handle a lot of work. This is due to work culture factors applied by Chinese migrant workers. Where workers are accustomed to uncertain work situations and work under pressure
Improvement of Business Process Modeling in Internal Audit Planning at an Indonesian State-Owned Enterprise
Get PDFThe previous research reported that there were gaps between the procedure of internal auditing at an Indonesian state-owned enterprise namely J Corp with the standard set by an international association of internal auditor. In order to harmonize the company\u27s procedure with the international standard for professional practice of internal auditing, this study analyzes the current business process of internal audit planning activity and proposes the to-be business process modeling. Special attention is given to the need of internal auditors as the key users of this activity to making improvement and adjustment the future business process model. This study applied qualitative approach by conducting focus group discussion with the internal auditors, reviewing the standard operating procedure of internal auditing, and observing the walkthrough of internal audit planning that have been done by internal auditors. By using Business Process Modeling and Notation, this study will draw graphical notation to depict steps in business process of company\u27s internal audit planning. The result of study is immediately beneficial for the internal audit function at the J Corp to improve their standard operating procedure by conforming to the international standard and also enriches the literature for practitioners to know-how to improve the business process in the internal audit planning
Smart Water Control Berbasis Arduino dengan Integrasi Motor Stepper dan Sensor Ultrasonik
No full textPemborosan air akibat kelalaian mematikan keran sering terjadi di rumah tangga, asrama, dan fasilitas umum. Air yang meluap tidak hanya boros tetapi juga dapat merusak properti. Beberapa penelitian sebelumnya menggunakan sensor level air atau solenoid valve, namun masih memiliki keterbatasan presisi dan fleksibilitas, terutama untuk keran air konvensional. Penelitian ini mengusulkan Smart Water Control berbasis Arduino dengan motor stepper dan sensor ultrasonik untuk solusi lebih presisi dan adaptif. Sistem ini mengintegrasikan motor stepper EM-93 03611 STH-39D172, Arduino UNO, dan sensor Ultrasonik SRF04 untuk memantau ketinggian air. Arduino mengendalikan motor stepper guna memutar keran 90° sebanyak 50 langkah per-siklus secara otomatis dengan buka-tutup berdasarkan data sensor. Kebaruan dari sistem ini terletak pada penggunaan motor stepper untuk keran mekanis konvensional yang memungkinkan kontrol presisi tanpa modifikasi plumbing.Pemborosan air akibat kelalaian dalam mematikan keran bak mandi masih menjadi permasalahan umum. Untuk mengatasi hal tersebut, dikembangkan sistem Smart Water Control berbasis Arduino yang mengintegrasikan motor stepper dan sensor ultrasonik. Sistem ini menggunakan motor Stepper Variable Reluctance Unipolar tipe EM-93 03611 STH-39D172, Arduino Uno sebagai mikrokontroler, serta sensor ultrasonik SRF04 untuk mendeteksi ketinggian air dalam bak mandi. Motor stepper dengan 200 langkah per putaran dan kecepatan 80 rpm dikendalikan melalui sinyal digital dari Arduino. Berdasarkan pembacaan sensor, sistem secara otomatis memutar keran 90° searah jarum jam untuk membuka dan 90° berlawanan arah jarum jam untuk menutup, dengan total 50 langkah per siklus. Sistem ini menawarkan solusi otomatis yang cerdas dan efisien untuk mengontrol aliran air, sehingga dapat mencegah pemborosan dan meningkatkan efisiensi penggunaan konsumsi air
Perancangan Trajectory Planning pada Case packer Delta Robot 3-Axis dengan Interpolasi Polinomial Kubik untuk Meningkatkan Pick Rate Produk
No full textMekanisme pick and place dinamis untuk pemindahan objek umumnya menggunakan robot paralel, seperti Delta robot 3-axis, yang membutuhkan koordinasi sistem yang teratur dan bekerja dengan cepat. Objek dijalankan melalui konveyor infeed secara kontinyu sehingga robot perlu bergerak dengan cepat dan tepat agar tidak ada produk yang terlewat. Dengan trajectory planning robot, resiko produk tidak terambil dapat diminimalkan. Dalam suatu kasus di sebuah industri, selama 10 menit hanya memiliki pick rate 82%, masih di bawah target standar, 90%. Dalam artikel ini dijelaskan implementasi metode interpolasi polinomial kubik dan inverse kinematic untuk membuat trajectory planning. Pengembangan metode baru dengan modifikasi gerakan robot lebih efisien dibandingkan dengan metode sebelumnya. Rangkaian gerakan robot awal dipangkas dari 5 menjadi 4 gerakan, menghasilkan lintasan yang lebih pendek dan mengurangi waktu bergerak yang tidak diperlukan. Dengan penggunaan metode tersebut didapatkan hasil simulasi pick rate produk sebesar 91,7% dan sudah melampaui target yang ditentukan.Mekanisme pick and place dinamis untuk pemindahan objek umumnya menggunakan robot paralel, salah satunya Delta robot 3-axis. Penerapan Pick and Place dinamis membutuhkan koordinasi sistem yang teratur dan bekerja dengan cepat. Objek berupa produk yang akan dikemas dijalankan melalui konveyor infeed secara kontinyu sehingga robot perlu bergerak dengan cepat dan tepat agar dapat mengambil produk tanpa terlewat. Trajectory planning pada robot menjadi begitu penting karena gerakan yang tidak efisien akan menyebabkan adanya risiko produk tidak terambil oleh mesin. Dalam suatu kasus di sebuah industri, didapatkan data dari pengamatan selama 10 menit hanya memiliki pick rate 82%, dimana masih di bawah target standar yaitu lebih dari 90%. Dalam artikel ini dijelaskan implementasi metode polinomial kubik dan inverse kinematic untuk membuat trajectory planning yang efisien dan dapat meningkatkan pick rate dari robot. Pengembangan trajectory planning baru yang menggunakan rangkaian gerakan robot modifikasi lebih efisien dibandingkan dengan penggunaan rangkaian gerakan robot referensi. Rangkaian gerakan robot referensi dipangkas dari 5 gerakan menjadi 4 gerakan, menghasilkan lintasan yang lebih pendek dan mengurangi waktu bergerak yang tidak diperlukan. Dengan penggunaan trajectory planning tersebut didapatkan hasil simulasi pick rate produk sebesar 91,7% dan sudah melampaui target yang ditentukan
Development of ViScan: A Mobile Application for Skin Cancer Detection Using Ionic Framework and YOLOv10x
No full textSkin cancer is a common global health issue, with the number of cases continuing to rise worldwide. Early detection is crucial for improving patient outcomes, but traditional detection methods often require significant time, cost, and medical expertise. To address this challenge, this research focuses on developing a mobile application that leverages deep learning, specifically the YOLOv10x model, to enable fast and accurate detection of skin lesions. This application aims to provide an easy-to-use platform for self-monitoring skin health, particularly for individuals in remote areas with limited access to medical facilities. The system uses the HAM10000 dataset, which consists of a diverse collection of dermoscopy images of skin lesions, to train the YOLOv10x object detection model for real-time detection on mobile devices. By leveraging TensorFlow.js and Node.js, the model processes skin images and provides real-time results with precision and efficiency. The mobile application, developed using the Ionic Framework, ensures cross-platform compatibility and a responsive, intuitive user interface. System performance was evaluated using key metrics such as Precision (84.2%), Recall (86.3%), mAP (89.2%), and F1 Score (85.2%), demonstrating its effectiveness in early skin cancer detection. The potential of this application extends beyond detection, contributing to society by raising awareness and offering an accessible, low-cost screening solution
Development of AI-Based Public Safety System with Face Recognition Using CNN and SVM Models in Real-Time
No full textSexual crimes are an increasing problem, with many cases difficult to identify due to the limitations of existing surveillance systems. This study aims to develop an Artificial Intelligence (AI)-based system using Convolutional Neural Network (CNN) and Support Vector Machine (SVM) for gender identification in order to support sexual crime investigations. The methods used include processing facial image datasets, training models using CNN for feature extraction, and SVM for gender classification. The results showed that the CNN model achieved an accuracy of 90.15%, while the SVM model only achieved an accuracy of 82.16%. Further evaluation with a confusion matrix showed that CNN was more accurate in classifying gender than SVM. With these results, the developed system has the potential to help authorities identify perpetrators of sexual crimes more quickly and accurately. The dataset used consists of 23,706 grayscale facial images of 48x48 pixels, with a balanced distribution of male and female samples. The CNN architecture includes three convolutional blocks and achieves 90.15% accuracy. Although designed for real-time operation, inference speed needs further validation using FPS or latency metrics on specific hardware platforms.Sexual crimes are an increasing problem, with many cases difficult to identify due to the limitations of existing surveillance systems. This study aims to develop an Artificial Intelligence (AI)-based system using Convolutional Neural Network (CNN) and Support Vector Machine (SVM) for gender identification in order to support sexual crime investigations. The methods used include processing facial image datasets, training models using CNN for feature extraction, and SVM for gender classification. The results showed that the CNN model achieved an accuracy of 90.15%, while the SVM model only achieved an accuracy of 82.16%. Further evaluation with a confusion matrix showed that CNN was more accurate in classifying gender than SVM. With these results, the developed system has the potential to help authorities identify perpetrators of sexual crimes more quickly and accurately. The dataset used consists of 23,706 grayscale facial images of 48x48 pixels, with a balanced distribution of male and female samples. The CNN architecture includes three convolutional blocks and achieves 90.15% accuracy. Although designed for real-time operation, inference speed needs further validation using FPS or latency metrics on specific hardware platforms
A Real-Time Hand Gesture Control of a Quadcopter Swarm Implemented in the Gazebo Simulation Environment
No full textWith the advancement of technology, human-robot interaction (HRI) is becoming more intuitive, including through hand gesture-based control. This study aims to develop a real-time hand gesture recognition system to control a quadcopter swarm within a simulated environment using ROS and Gazebo. The system utilizes Google\u27s MediaPipe framework for detecting 21 hand landmarks, which are then processed through a custom-trained neural network to classify 13 predefined gestures. Each gesture corresponds to a specific command such as basic motion, rotation, or swarm formation, and is published to the /cmd_vel topic using the ROS communication framework. Simulation tests were performed in Gazebo and covered both individual drone maneuvers and simple swarm formations. The results demonstrated a gesture classification accuracy of 90%, low latency, and stable response across multiple drones. This approach offers a scalable and efficient solution for real-time swarm control based on hand gestures, contributing to future applications in human-drone interaction systems
Performance Evaluation of a Palm Oil Factory Using the Balanced Scorecard: A Case Study Approach
No full textThe purpose of this study is to evaluate the performance of a palm oil mill using the Balanced Scorecard (BSC) model at PT. Perkebunan Nusantara IV Pasir Mandoge. This research employs a quantitative approach, utilizing numerical data and statistics to answer the research questions. The data consists of primary data collected through interviews and secondary data from financial reports at PTPN IV. The data were analyzed using the Balanced Scorecard concept, which divides performance measurements into financial and non-financial perspectives. The results of the study indicate that the financial perspective yields good performance results, which need to be maintained by increasing income through marketing efforts. This can be achieved by enhancing product innovation, improving efficiency on each production machine, and developing facilities and services to boost income, maximizing existing resources for optimal returns. From the customer\u27s perspective, performance results are poor due to inadequate attention to customer complaints and feedback, as the company focuses more on growing its market share, rather than reducing complaints through customer retention and acquisition strategies. Recommendations for the customer perspective include minimizing defective products to ensure quality, maintaining strong customer relationships, and offering appealing incentives. From an internal business process perspective, improvements in company performance are evident, as demonstrated by faster processing times, which in turn lead to higher customer satisfaction. The operational perspective shows satisfactory results, as seen in monthly outputs compared to company averages. The learning and growth perspective indicates even better outcomes. An increase in the number of employees participating in training and development has a positive impact on employee performance, satisfaction, and productivity, ultimately benefiting the company in the long run