SASon Ali Akbar

Son Ali Akbar

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Son Ali Akbar has a total of 45 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    AI-driven stress monitoring in melon crops via graph neural networks

    by Son Ali Akbar, Jihad Rahmawan, Etika Dyah Puspitasari, Anton Yudhana, Novi Febrianti

    Melon cultivation is highly vulnerable to abiotic and biotic stress, and early detection remains difficult when monitoring relies on a single sensing modality. This study investigated a multimodal stress-classification framework that combined root-zone measurements and canopy reflectance descriptors for melon monitoring under greenhouse conditions. Soil pH, nitrogen, phosphorus, potassium, and temperature were acquired using an RS485 multi-parameter sensor, while canopy images were captured using a Raspberry Pi NoIR camera and converted into Normalized Difference Vegetation Index features. Each synchronized observation was represented as a graph with fixed variable nodes and correlation-based edges, enabling relation-aware learning through a Graph Convolutional Network. The proposed model was evaluated using cross-validation and compared against conventional machine learning and non-graph deep learning baselines. The graph-based model achieved the best overall classification performance, indicating that explicit modeling of soil-canopy dependencies improved discrimination between healthy and stressed plants. The results suggest that graph-structured multimodal fusion is a promising strategy for AI-assisted crop stress monitoring.

  • Open Access

    A Deep Learning Approach for Gram-Stained Bacterial Classification Using an Enhanced Hybrid Deep Convolutional Neural Network

    by Kamarul Hawari Ghazali, Son Ali Akbar, Razali Muda, Sunardi Sunardi, Phisca Rosyady, Murein Mardhi, M.S. Najib, Ji Yuanfa

    Accurate identification of bacteria in Gram-stained microscopy images is essential for timely clinical decisions. Manual inspection is time-consuming, error-prone, and often fails to differentiate morphologically similar species. This study proposes an Enhanced Hybrid Deep Convolutional Neural Network (EHDCNN) for automated classification of six types of Gram-stained bacteria. The model combines multi-scale deep feature extraction via a fine-tuned ResNet101 with an Error-Correcting Output Codes (ECOC) Support Vector Machine (SVM) classifier to improve robustness and generalization. An augmented dataset was used to address class imbalance and enhance feature diversity. The EHDCNN achieved an average accuracy of 99.5%, with high recall, specificity, and precision—outperforming its baseline (HDCNN) and classical models including Naïve Bayes, MLP, and SVM. These results support the potential of EHDCNN as a reliable tool for bacterial classification in clinical diagnostic workflows.

  • Open Access

    Deteksi Dini Banjir Menggunakan Algoritma Deteksi Tepi Pada Citra Sungai Pulu, Sulawesi Tengah

    by Leonard Onsik, Son Ali Akbar, Kartika Firdausy

    Banjir merupakan salah satu bencana alam yang sering terjadi di berbagai wilayah di Indonesia dan dapat mengakibatkan kerusakan lingkungan, infrastruktur, serta mengancam keselamatan jiwa masyarakat. Kabupaten Sigi di Provinsi Sulawesi Tengah merupakan salah satu wilayah rawan banjir yang memerlukan sistem pemantauan dan deteksi dini berbasis teknologi. Penelitian ini bertujuan untuk merancang dan mengimplementasikan sistem deteksi banjir berbasis pengolahan citra digital menggunakan kamera CCTV dan Raspberry Pi, serta melakukan perbandingan akurasi tiga metode deteksi tepi yaitu Canny, Sobel, dan Prewitt. Akuisisi citra dilakukan secara real-time pada aliran Sungai Pulu, dan sistem memproses citra untuk mendeteksi permukaan air menggunakan segmentasi warna HSV dan metode deteksi tepi. Hasil pengujian menunjukkan bahwa metode Canny memiliki akurasi pembacaan tertinggi dengan rata-rata 98,36%, diikuti Sobel dengan 98,18%, dan Prewitt dengan 85,51%. Sistem ini menunjukkan kinerja yang stabil bahkan pada kondisi pencahayaan rendah. Dengan demikian, sistem ini dapat dijadikan solusi efektif dan berbiaya rendah sebagai alat bantu peringatan dini banjir di daerah rawan seperti Kabupaten Sigi.

  • Open Access

    Integrasi Sensor Pintar dengan Sistem IoT untuk Memantau Kadar Gas Nitrogen pada Reaktor Biogas

    by Son Ali Akbar, Muhammad Azis Sidiq, Anton Yudhana, Rahmat Rian Hidayat

    Energi terbarukan merupakan salah satu energi yang berasal dari proses alam akan tetapi tetap ada dan dapat digunakan secara terus menerus. Penggunaan energi baru terbarukan merupakan bentuk masyarakat mengurangi penggunaan bahan bakar fosil. Salah satu energi baru terbarukan yang berada pada lingkup rumah tangga yaitu biogas. Biogas adalah gas yang dihasilkan oleh mikroorganisme selama fermentasi anaerobik bahan organik. Penelitian ini mengusulkan metode monitoring gas nitrogen berbasis Internet of Things (IoT) yang mampu mengetahui kadar gas nitrogen yang terkandung dalam biogas. Perancangan perangkat keras sistem untuk mendeteksi kadar gas nitogen digunakan perangkat sensor MQ-135 dan ESP8266 untuk pengolahan data melalui internet atau secara langsung. Hasil monitoring menunjukkan kadar gas nitrogen yang terdapat pada biogas tidak melebihi 20000 ppm. Sesuai dengan kadar standar yang ditetapkan berupa kadar nitrogen sebesar 2% - 5% atau 20.000 ppm. Oleh karena itu, biogas memenuhi kriteria.

  • Open Access

    PID Control Tuning Based on Wind Speed Sensor in Flying Robot

    by Fadlur Rahman T Hasan, Son Ali Akbar

    The problem that is often faced by flying robots when carrying out the Vertical Take Off Landing (VTOL) process is the lack of stability of the vehicle due to differences in wind speed at any time. This is because the PID that has been pre-tuned is the PID at a certain wind speed and it is possible that during the race the wind speed suddenly changes, causing the vehicle to be less stable in carrying out the mission. Therefore, this study proposes a PID Control Tuning Control system based on the Wind Speed Sensor. In experiments that have been carried out with anemometer readings of 1-5 m/s, the ideal tuning results are obtained with each parameter P_roll = 0.1453, I_roll = 0.0892, D_roll = 0.004, P_pitch = 0.144, I_pitch = 0.09, D_pitch = 0.004, P_yaw = 0.184, I_yaw = 0.0184, D_yaw = 0.00309. In experiments with anemometer readings of 6-10 m/s, the ideal tuning results were obtained with each parameter P_roll = 0.148, I_roll = 0.0905, D_roll = 0.004, P_pitch = 0.1444, I_pitch = 0.09, D_pitch = 0.004, P_yaw = 0.1867, I_yaw = 0.0181, D_yaw = 0.0037. In an experiment with an anemometer reading of 11-15 m/s, the ideal tuning results were obtained with each parameter P_roll = 0.1494, I_roll = 0.09, D_roll = 0.004, P_pitch = 0.1457, I_pitch = 0.0902, D_pitch = 0.004, P_yaw = 0.1894, I_yaw = 0.018, D_yaw = 0.0037. PID adjustment based on this anemometer sensor utilizes the latest real-time wind speed data to support the robot in order to overcome instability in certain wind conditions by tuning PID so that the vehicle can maintain stability when carrying out certain missions.


  • Open Access

    Analysis of the Influence of Number of Segments on Similarity Level in Wound Image Segmentation Using K-Means Clustering Algorithm

    by Furizal Furizal, Syifa’aH Setya Mawarni, Son Ali Akbar, Anton Yudhana, Murinto Kusno

    This study underscores the importance of wound image segmentation in the medical world to speed up first aid for victims and increase the efficiency of medical personnel in providing appropriate treatment. Although the body has a protective function from external threats, the skin can be easily damaged and cause injuries that require rapid detection and treatment. This study used the K-Means clustering algorithm to segment the external wound image dataset consisting of three types of wounds, namely abrasion, puncture, and laceration. The results showed that K-Means clustering is an effective method for segmenting wound images. The greater the number of segments used, the better the quality of the resulting segmentation. However, it is necessary to take into account the specific characteristics of each type of wound and the number of segments used in order to choose the most suitable segmentation method. Evaluation using various metrics, such as VOI, GCE, MSE, and PSNR, provides an objective assessment of the quality of segmentation. The results showed that abrasion wounds were easier to segment compared to puncture wounds and lacerations. In addition, the size of the image file also affects the speed of program execution, although it is not always absolute depending on the characteristics of the image.


  • Open Access

    Oil Palm USB (Unstripped Bunch) Detector Trained on Synthetic Images Generated by PGGAN

    by Wahyu Sapto Aji, Kamarul Hawari Bin Ghazali, Son Ali Akbar

    Identifying Unstriped Bunches (USB) is a pivotal challenge in palm oil production, contributing to reduced mill efficiency. Existing manual detection methods are proven time-consuming and prone to inaccuracies. Therefore, we propose an innovative solution harnessing computer vision technology. Specifically, we leverage the Faster R-CNN (Region-based Convolution Neural Network), a robust object detection algorithm, and complement it with Progressive Growing Generative Adversarial Networks (PGGAN) for synthetic image generation. Nevertheless, a scarcity of authentic USB images may hinder the application of Faster R-CNN. Herein, PGGAN is assumed to be pivotal in generating synthetic images of Empty Fruit Bunches (EFB) and USB. Our approach pairs synthetic images with authentic ones to train the Faster R-CNN. The VGG16 feature generator serves as the architectural backbone, fostering enhanced learning. According to our experimental results, USB detectors that were trained solely with authentic images resulted in an accuracy of 77.1%, which highlights the potential of this methodology. However, employing solely synthetic images leads to a slightly reduced accuracy of 75.3%. Strikingly, the fusion of authentic and synthetic images in a balanced ratio of 1:1 fuels a remarkable accuracy surge to 87.9%, signifying a 10.1% improvement. This innovative amalgamation underscores the potential of synthetic data augmentation in refining detection systems. By amalgamating authentic and synthetic data, we unlock a novel dimension of accuracy in USB detection, which was previously unattainable. This contribution holds significant implications for the industry, ensuring further exploration into advanced data synthesis techniques and refining detection models.


  • Open Access

    The Stability of the Fixed Wings Unmanned Aircraft Using the PID Method at KRTI 2019

    by Pramudya Wisnujati, Son Ali Akbar

    This study designed a UAV with an unmanned aircraft type with a Fixed Wings model that can perform the PID tunning process manually using the FBWA flight mode. An unmanned aircraft with the Fixed Wings type is designed to perform PID Roll tuning based on altitude and flight speed using the trial and error method increases the Proll value of 0.06 x 2 until the oscillation disappears and the final value of Proll is reduced by 50%. For the Iroll value, the increase in the value made is 0.1 x 2, increase it until it is sufficient and the final value of Iroll is reduced by 50%. For the Droll value, increase the value of 0.004 x 2 until the oscillation disappears and the final value of the Droll is reduced by 50%.


  • Open Access

    Early Bacterial Detection in Bloodstream Infection using Deep Transfer Learning Algorithm

    by Son Ali Akbar, Kamarul Hawari Ghazali, Habsah Hasan, Wahyu Sapto Aji, Anton Yudhana

    An infection caused by bacteria can lead to severe complications affecting bloodstream disease. At present, blood cultures are used to identify bacteria. However, blood culture is a time-consuming and labor-intensive method of diagnosing disease. The effect of delayed early diagnosis is that it influences the mortality risk. Thus, it is urgent to develop an initial prediction model to identify patients with bloodstream infections. This paper focused on classifying the bacteria using a deep learning approach. Besides, techniques of deep learning have the ability to enhance the bacterial classification process more effectively. Using the transfer learning-based convolutional neural network technique involved to develop our model. In addition, we compared the proposed model with another model used to find the best results. Compared to other models, the proposed model achieved an evaluation score with high accuracy of 98.62%. Medical decision-making may benefit from the proposed approach.


  • Open Access

    Design of One-Phase Inverter Using EGS002 with SPWM

    by Nisfa Sri Ayu Kaliky, Son Ali Akbar, Ahmad Raditya Cahya Baswara

    The more advanced the development of the technological world in Indonesia, the needs of the community in the use of electrical energy are also higher so that an alternative tool is needed that can produce a voltage source such as PLN. In this thesis, a 1-phase inverter will be designed that can produce output in the form of an AC voltage of 220 volts with a frequency of 50 Hz and has a good feedback voltage so that the inverter output voltage can be stable at an AC voltage value of 220 volts. The 1-phase inverter is made using a full bridge inverter circuit, the SPWM (sinusoidal pulse width modulation) signal is generated using EGS002 by ASIC (Application Specific Integreted Circuit), and the voltage correction transformer used has a maximum working power of 600 watts. According to the results of the no-load test the 1-phase inverter can produce an AC voltage on the output side of 220 to 230 volts with 50 Hz frequencys and the results of the load test show that the inverter has a good feedback voltage so that even when the input voltage drops, the inverter output voltage value is still above 220 volts.