Claim Missing Document
Check
Articles

Designing stair climbing wheelchairs with surface prediction using theoretical analysis and machine learning Chawaphan, Pharan; Maneetham, Dechrit; Crisnapati, Padma Nyoman
Indonesian Journal of Electrical Engineering and Computer Science Vol 38, No 1: April 2025
Publisher : Institute of Advanced Engineering and Science

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.11591/ijeecs.v38.i1.pp120-132

Abstract

Urban settings present considerable obstacles for those use personal mobility wheelchairs, especially when it comes to manoeuvring stairs. The objective of this study is to improve the safety and ease of use of wheelchairs designed for ascending stairs. The study aims to tackle the significant issue of instability and limited ability to adjust to different types of terrain. This research employs a holistic methodology that combines theoretical dynamic analysis, hardware design and simulation, and field testing, in addition to advanced machine learning approaches for surface prediction. Theoretical models guarantee the stability of the wheelchair, while hardware simulations offer valuable insights into its structural integrity. The data obtained from inertial measurement unit (IMU) sensors during field tests is analysed and categorised using models like random forest and gradient boosting, which exhibit exceptional accuracy in forecasting movement circumstances. The results demonstrate that the implementation of these combined techniques greatly enhances the wheelchair’s capacity to safely manoeuvre over urban barriers. The study finds that the suggested solutions show great potential for creating intelligent mobility aids, which might be used to improve accessibility for those with mobility limitations.
Hybrid object detection and distance measurement for precision agriculture: integrating YOLOv8 with rice field sidewalk detection algorithm Tungkasthan, Anucha; Jongsawat, Nipat; Crisnapati, Padma Nyoman; Thwe, Yamin
IAES International Journal of Artificial Intelligence (IJ-AI) Vol 14, No 2: April 2025
Publisher : Institute of Advanced Engineering and Science

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.11591/ijai.v14.i2.pp1507-1517

Abstract

This study aims to propose a new approach to semantic segmentation of sidewalk images in rice fields using the YOLOv8 algorithm, with the objective of enhancing agricultural monitoring and analysis. The experimental process involved preparing the development environment, extracting data from JSON, and training the model using YOLOv8. Evaluation reveals consistent and accurate sidewalk detection with a confidence score of 0.9-1.0 across various environmental conditions. Confusion matrix and precision-recall analysis confirmed the robustness and accuracy of the model. These findings validate the effectiveness of the approach in detecting and measuring sidewalks with high precision, potentially improving agricultural monitoring. The novelty of this study lies in the utilization of the RIFIS-D algorithm as an integral part of a hybrid approach with YOLOv8. This hybridization enriches the model with additional capability to detect the distance between the sidewalk and the tractor, addressing specific needs in agricultural applications. This contribution is significant in the advancement of automatic navigation and monitoring technology in agriculture, enabling the implementation of more sophisticated and efficient systems in field operations.
Evaluating the Effectiveness of Alzheimer’s Detection Using GANs and Deep Convolutional Neural Networks (DCNNs) Pamungkas, Yuri; Syaifudin, Achmad; Crisnapati, Padma Nyoman; Hashim, Uda
International Journal of Robotics and Control Systems Vol 5, No 2 (2025)
Publisher : Association for Scientific Computing Electronics and Engineering (ASCEE)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.31763/ijrcs.v5i2.1855

Abstract

Alzheimer’s is a gradually worsening condition that damages the brain, making timely and precise diagnosis essential for better patient care and outcomes. However, existing detection methods using DCNNs are often hampered by the problem of class imbalance in datasets, particularly OASIS and ADNI, where some classes are underrepresented. This study proposes a novel approach integrating GANs with DCNNs to tackle class imbalance by creating synthetic samples for underrepresented categories. The primary focus of this research is demonstrating that using GANs for data augmentation can significantly strengthen DCNNs performance in Alzheimer's detection by balancing the data distribution across all classes. The proposed method involves training DCNNs with both original and GAN-generated data, with data partitioning of 80:10:10 for training/ validation/ testing. GANs are applied to generate new samples for underrepresented classes within the OASIS and ADNI datasets, ensuring balanced datasets for model training. The experimental results show that using GANs improves classification performance significantly. In the case of the OASIS dataset, the mean accuracy and F1 Score rose from 99.64% and 95.07% (without GANs) to 99.98% and 99.96% (with GANs). For the ADNI dataset, the average accuracy and F1 Score improved from 96.21% and 93.01% to 99.51% and 99.03% after applying GANs. Compared to existing methods, the proposed GANs + DCNNs model achieves higher accuracy and robustness in detecting various stages of Alzheimer's disease, particularly for minority classes. These findings confirm the effectiveness of GANs in improving DCNNs' performance for Alzheimer's detection, providing a promising framework for future diagnostic implementations.
Analisis Komparatif Model Pembelajaran Mesin untuk Klasifikasi Biner pada Data HVAC Gusti Ngurah Putra Arimbawa, I; Ayu Juli Astari, Ni Made; Crisnapati, Padma Nyoman; Devi Novayanti, Putu; Duika Adi Sucipta, I Kadek; Panji Anggara, Dicky; Yuda Danuarta, I Putu
Jurnal Algoritma Vol 22 No 2 (2025): Jurnal Algoritma
Publisher : Institut Teknologi Garut

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.33364/algoritma/v.22-2.2657

Abstract

The HVAC (Heating, Ventilation, and Air Conditioning) system plays a crucial role in maintaining thermal comfort and energy efficiency in commercial and industrial buildings. However, early detection of anomalies or failures in this system is often suboptimal, leading to increased energy consumption, reduced operational performance, and high maintenance costs. This study aims to develop and evaluate various machine learning models for real-time anomaly detection in HVAC systems, using a real-world dataset that includes 11 key operational variables. Several algorithms are used, including Logistic Regression, K-Nearest Neighbors (KNN), Support Vector Classifier (SVC), Decision Tree, Random Forest, AdaBoost, Gradient Boosting, and XGBoost. The dataset is labeled based on dynamic deviations between actual temperature and setpoint using the Exponential Moving Average (EMA) approach, which allows for adaptive anomaly labeling. The experimental results show that the XGBClassifier achieves an accuracy of 99.32%, with precision and recall of 0.98 each, and an F1-score of 0.98, making it the best model for detecting anomalies in a balanced manner. Logistic Regression (accuracy 99.54%, F1-score 0.99) and Random Forest (accuracy 98.70%, F1-score 0.96) also proved to be reliable and computationally efficient alternatives. Thus, this research not only provides a comprehensive comparison of models but also emphasizes the novelty of the adaptive labeling strategy to support real-time anomaly detection in HVAC systems, which can enhance energy efficiency while reducing maintenance costs.
LQ45 Index Stock Market Prediction: A Deep Learning Approach using LSTM with Bayesian Optimization Crisnapati, Padma Nyoman; Putu Devi Novayanti; Dian Pramana
Jurnal Sistem Komputer dan Informatika (JSON) Vol. 7 No. 2 (2025): Desember 2025
Publisher : Universitas Budi Darma

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30865/json.v7i2.8925

Abstract

This study investigates the application of Long Short-Term Memory (LSTM) models with Bayesian Optimization for predicting stock price movements in the LQ45 Index, a collection of the 45 most liquid stocks on the Indonesia Stock Exchange. The primary objective is to enhance prediction accuracy by addressing the challenges of volatile stock markets and inefficient hyperparameter tuning. Historical data, including daily closing prices from January 2020 to October 2024, was processed using Min-Max Scaling and transformed into time-series input features with a 60-time-step window. Bayesian Optimization was employed to fine-tune key hyperparameters such as LSTM units, dropout rate, and learning rate, optimizing the model's performance. The results revealed that the LSTM model accurately captured trends for stocks with stable price patterns, such as ACES, ASII, and MTEL, achieving low Mean Absolute Percentage Error (MAPE) and Root Mean Square Percentage Error (RMSPE). However, stocks with high volatility, like AMMN and ITMG, exhibited higher prediction errors, indicating limitations in modeling complex patterns. The study highlights that while LSTM with Bayesian Optimization is highly effective for stable stocks, additional preprocessing and advanced modeling techniques are required for volatile stocks. This research demonstrates the potential of machine learning in supporting stock market decision-making, contributing to the development of more robust and efficient financial prediction tools for investors navigating dynamic markets.
PID-Controlled Gyroscopic Stabilization for Roll Balancing: A Simulation and Experimental Study Crisnapati, Padma Nyoman; Putu Devi Novayanti; Ni Ketut Ira Permata Adi; I Putu Agung Mas Aditya Warman; Anak Agung Istri Cintya Prabandari; I Made Darma Susila
Jurnal Sistem Komputer dan Informatika (JSON) Vol. 7 No. 2 (2025): Desember 2025
Publisher : Universitas Budi Darma

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30865/json.v7i2.9001

Abstract

In marine engineering, stabilizing boat roll motion under wave-induced disturbances is a crucial problem where traditional approaches frequently have drawbacks in terms of responsiveness, energy efficiency, and adaptability. In this study, a PID-controlled gyroscopic stabilization system for boat roll balancing is designed, simulated, and experimentally validated. To capture the coupled dynamics of servo motor behavior, gyroscopic torque generation, and boat roll motion, a thorough dynamic model was created. Four gain configurations—Low, Moderate, High, and Very High—were evaluated using the model-guided PID parameter tuning that was implemented in Python. The mechanical system incorporates a gyroscopic flywheel driven by BLDC and mounted on a servo-controlled cradle. An ESP32 microcontroller processes real-time roll angle feedback from an MPU6050 sensor. According to simulation results, the ideal balance between rise time (~300 ms), overshoot (~2°), and settling time (~1 s) was reached with moderate PID gains. While High and Very High gains displayed instability because of unmodeled vibrations and sensor noise, a scaled physical prototype that was built and tested under controlled disturbances demonstrated strong alignment with simulation trends for Low and Moderate gains. The results show that moderate gains, which offer both quick stabilization and reliable performance, offer the most useful configuration for real-world applications. This work contributes a validated methodology for optimizing PID-controlled systems in dynamic environments by bridging the gap between theoretical modeling and practical implementation of marine gyroscopic stabilization.
Efficient YOLO-based models for real-time ceramic crack detection Maungmeesri, Benchalak; Khonthon, Sasithorn; Maneetham, Dechrit; Crisnapati, Padma Nyoman
IAES International Journal of Artificial Intelligence (IJ-AI) Vol 15, No 1: February 2026
Publisher : Institute of Advanced Engineering and Science

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.11591/ijai.v15.i1.pp852-860

Abstract

The following research work systematically compares four variants of you only look once (YOLO), namely, YOLOv8, YOLOv9, YOLOv10, and YOLOv11 proposed recently, considering the key properties required to perform ceramic surface crack detection tasks with high computational efficiency, real-time inference speed, and low memory usage. A total of 300 images of ceramic surfaces were collected with manually labeled cracks and divided into training, validation, and testing sets in portions of 263, 22, and 15 images, respectively. Each of the four YOLO variants was trained for 50 and 100 epochs, and each was evaluated regarding mean average precision (mAP), inference time, model size, and computational complexity in giga floating point operations per second (GFLOPs). YOLOv9 produced the highest accuracy with mAP values as high as 0.752-0.79 but the highest cost in terms of increased computational complexity. However, among these methods, YOLOv8 can produce the fastest inference (~2-2.3 ms) with a small memory footprint (~6 MB) with an acceptable accuracy of ~0.65-0.67. The results showed that YOLOv8 is the most feasible to deploy in resource constrained industrial automation environments. By offering this comparative study, the research attempts to provide hints for the selection of appropriate YOLO-based models by practitioners in quality control applications related to ceramic manufacturing.
Co-Authors Achmad Syaifudin Ade Widiyantara, I Putu Adi Yoga Dewantara, I Made Agus Sutrisna, I Kadek Anak Agung Istri Cintya Prabandari Aprilia Yustika Dewi, Aprilia Ardipa, Gede Sukra Arief Hadi Prasetyo Arief Hadi Prasetyo, Arief Hadi Arisandi, Ni Made Desi Aryasih, Putu Putri Ayu Juli Astari, Ni Made Ayu Nirma Lestari, Gusti Bunga Anindya, Made Cahyani, Agung Ayu Hanna Chawaphan, Pharan Dian Pramana Duika Adi Sucipta, I Kadek Dwi Suparyanta, Kadek Gusti Ngurah Putra Arimbawa, I Hanna Cahyani, Agung Ayu Haryantara, I Nyoman Hashim, Uda I Gede Mahendra Darmawiguna I Gusti Ayu Sri Melati, I Gusti Ayu I Ketut Resika Arthana I Komang Agus Ady Aryanto I Komang Ariesta Ananta, I Komang Ariesta I Komang Try Adi Stanaya, I Komang Try Adi I Made Darma Susila I Made Gede Sunarya I Nyoman Haryantara I Nyoman Kusuma Wardana I Putu Agung Mas Aditya Warman I Putu Dana Arista, I Putu Indriani, Ratri Dwi Jongsawat, Nipat Kadek Artawan Kadek Artawan, Kadek Khonthon, Sasithorn Lanang Nugraha, Made Lenny Natalia, Lenny M.Cs S.Kom I Made Agus Wirawan . Made Frans Aditya Bramantya Kusuma Made Frans Aditya Bramantya Kusuma, Made Frans Aditya Bramantya Made Windu Antara Kesiman Maemonah, Maemonah Maneetham, Dechrit Mariyantoni, I Kadek Yostab Maungmeesri, Benchalak Merta, I Gede Muchammad Naseer Ni Kadek Sumiari, Ni Kadek Ni Ketut Ira Permata Adi Ni Komang Oktari Permata Sari Oky Sanjaya, Kadek Pandu Wibawa S, I Wayan Panji Anggara, Dicky Prawira, Putu Yoka Angga Putra Yasa, Gede Agus Putu Angga Sudyatmika Putu Devi Novayanti Putu Putri Sanjani, Dewa Ayu Ratna Kartika Wiyati, Ratna Kartika Reditya Ary Prasetya, Agus Nyoman Rendy Syahrial, Lalu Ricky Aurelius Nutanto Diaz, Ricky Aurelius Roman Apriyansyah, Roman Shofwan Hanief Sri Darmaningsih, Luh Suarningsih, Putu Susena, I Komang Sutrisna Oka, Ketut Tri Thwe, Yamin Tirta Murdika, I Made Tungkasthan, Anucha Udayana, Ketut Wira Widiantara, Eka Putra Wikan Paramasila, Kadek Yoga Antara, I Made Yoka Angga Prawira, Putu Yostab Mariyantoni, I Kadek Yuda Danuarta, I Putu Yudiantara, I Made Yuri Pamungkas