Claim Missing Document
Check
Articles

Found 18 Documents
Search

Pemodelan Klasifikasi Penyakit Daun Tanaman Tomat dengan Convolutional Neural Network Algorithm Goklas Henry Agus Panjaitan; Simatupang, Frengki
KLIK: Kajian Ilmiah Informatika dan Komputer Vol. 4 No. 5 (2024): April 2024
Publisher : STMIK Budi Darma

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30865/klik.v4i5.1646

Abstract

Tomatoes constitute a horticultural crop with significant economic value and are extensively cultivated by the people of Indonesia, an agricultural country. Tomato productivity is anticipated to increase in 2021-2022. However, issues often arise during the plant care process when attempting to identify the type of disease affecting the leaves of tomato plants, owing to the nearly identical symptoms that make differentiation challenging. Tomato plant leaves play a crucial role in the plant growth process. Addressing the challenge of tomato leaf diseases necessitates a method that employs specific technology for easy detection. Before implementing this method, a classification model for tomato leaf plants must be developed. The goal of this modeling is to create an effective tomato leaf classification model for subsequent use in a disease detection system that utilizes IoT technology. The modeling of tomato plant disease classification involves the application of the Convolutional Neural Network algorithm. The modeling process encompasses data source analysis, data preprocessing, model formation, model training, model evaluation, and the subsequent acquisition of model results. The dataset utilized for implementation comprises 18,162 images of tomatoes. The accuracy of the training data surpasses that of the validation data, indicating the reliability of the data used for model development. The accuracy of the training data is recorded at 94.06%, signifying its suitability for modeling. Additionally, the loss function result on the training data is 7.8%, further affirming the quality of the training data for model development.
Analysis and Design of Regional Asset Leasing Management Information System in North Tapanuli Regency Sinaga, Ardiles; Simatupang, Frengki; Manalu, Istas; Saragi, Hadi Sutanto
TEPIAN Vol. 5 No. 4 (2024): December 2024
Publisher : Politeknik Pertanian Negeri Samarinda

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.51967/tepian.v5i4.3188

Abstract

Regional assets are something that has an essential role in the scope of local government because they have a complex value and are quite significant when compared to other components in the local government narrative. This requires local governments to be able to manage and utilize regional assets properly and correctly to increase Original Local Government Revenue in the region. Effective and efficient management of regional assets is one example of the management of the regional economy as shown in the Original Local Government Revenue. Improving the quality of regional asset management that is carried out professionally and modernly is a step taken to prioritize the principles of Good Governance. This research will discuss how to design a model for managing land assets of the North Tapanuli Regency local government that prioritizes the principles of Good Governance by utilizing information technology in the form of designing an information system that is expected to bring changes to the institutions that use it. The information system is expected to make the management of land assets spread across 15 sub-districts neater and more well-organized. In addition, the management of information related to asset leases can facilitate the reporting process and decision making. The purpose of this research is to design an information system as a reference for the Regional Finance and Asset Agency to develop an information system for managing regional assets and leasing regional assets by utilizing web-based and mobile technology. It is expected that through the information system designed in this research, the retribution payers can submit a Lease Application, upload the files of the Application Letter, and download the agreement letter through the application. Another technology that can be utilized in the information system to be built is the use of docker containers to run a lighter database and save time when storing all data related to the object of retribution, retribution payers, applications, lease agreement contracts, bills, and payments. This research also applies data storage technology that divides data into several separate units and then stored in one repository called object storage technology. This technology is used to store large data, create backup copies and archives, and store large amounts of images, graphics, video, or audio files.
Design and Implementation of Rain Gauge Measurement using Ultrasonic Sensor HC-SR04 (Case Study using Ombrometer Method) Simatupang, Frengki; Nainggolan, Waldo Owen
Piston: Journal of Technical Engineering Vol. 7 No. 2 (2024)
Publisher : Program Studi Teknik Mesin Universitas Pamulang

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32493/pjte.v7i2.36619

Abstract

There are several types of rainfall measurement method, such as tipping bucket, siphon, rain water weight and other various methods. This research developed the alternative and affordable method to measure the rain gauge using ultrasonic sensor. The method used to convert the volume of collected rain water to the level of a rain gauge is Ombrometer method. Ultrasonic Sensor HC-SR04 is chosen as a sensor to measure the water level in the collected rain water tank because it categorized as a low-cost sensor. Measuring the level of rainfall using an ultrasonic sensor is influenced by the design of the reservoir tank and also the resolution of the ultrasonic sensor. This research found that the difference the measurement using ultrasonic sensor between ombrometer formula is sufficiently linear with the difference average at 32.63%. The lookup table method is implemented to linearize the measurement based on the ombrometer value method. By using the linearized value, the design is capable to measure light rain category because the design has the sensitivity/resolution measurement of rain gauge on 0.48 mm. This research also found that the average value of the measurement using the ultrasonic sensor for each measured volume water injected to the tank using this design and use this value as a trigger to send the actual measurement result to the users. All parameters from this system are sent using Wi-Fi connectivity to the server and can be monitored using Online Web Dashboard.
Unreal Engine Based MCS Program Using Microcontroller Wemos D1 Mini, MPU-9250 Sensor and UDP Protocol Simatupang, Frengki; Manalu, Istas Pratomo
Piston: Journal of Technical Engineering Vol. 8 No. 1 (2024)
Publisher : Program Studi Teknik Mesin Universitas Pamulang

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32493/pjte.v8i1.40210

Abstract

The Motion Capture System (MCS) is a series of steps involving the capture, processing, and mapping of human, animal, or object movements into virtual characters. This process includes motion recording, data processing with computers, and mapping the results into digital models. This article discusses the use of MCS for realistically reconstructing and simulating behavior, which is valuable in various applications such as entertainment, animation, computer games, sports, and rehabilitation. MCS faces challenges in accurately capturing the diversity and complexity of body movements, requiring advanced hardware and software. Unreal Engine, a development platform known for its ability to produce high-quality graphics and realistic physics systems, is often used to integrate MCS. The MPU9250 sensor, a 9-axis inertial sensor combining an accelerometer, gyroscope, and magnetometer, is used in this project to detect linear motion, rotation, and orientation with high accuracy. This research aims to develop an MCS program based on Unreal Engine using the MPU9250 sensor and the UDP protocol, with the Wemos D1 Mini as the client module to process and transmit sensor data via Wi-Fi. This process involves collecting sensor data, transmitting data via UDP, and processing the data in Unreal Engine to control the movement of objects or characters. Sensor calibration is performed to obtain accurate reference values, and the collected data is used to calculate pitch, roll, and yaw values. This data is then sent to Unreal Engine to display animations corresponding to human movements. Testing shows that the generated movements align with those demonstrated, although there are minor acceptable errors. This research highlights the importance of combining sensor technology and software to produce an accurate and responsive motion control system, providing an interactive and realistic user experience.
Sistem Pengendalian dan Pemantauan Terpusat pada Perangkat IoT Terdistribusi Sinambela, Eka Stephani; Simatupang, Frengki; Wowiling, Gerry Italiano; Sigiro, Marojahan Mula Timbul; Manalu, Istas Pratomo; Silalahi, Sari Muthia; Siagian, Pandapotan
Piston: Journal of Technical Engineering Vol. 8 No. 1 (2024)
Publisher : Program Studi Teknik Mesin Universitas Pamulang

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32493/pjte.v8i1.48473

Abstract

Perkembangan pesat Internet of Things (IoT) telah memungkinkan berbagai aplikasi cerdas; namun, pengelolaan banyak perangkat IoT yang tersebar secara manual masih kurang efisien dan memakan waktu. Penelitian ini bertujuan untuk mengembangkan sistem pemantauan dan pengendalian terpusat untuk perangkat IoT yang tersebar dengan menggunakan arsitektur master-agent. Master berfungsi sebagai pusat kendali yang mengumpulkan data dari berbagai agent serta memungkinkan manajemen terpusat melalui API Gateway yang memfasilitasi komunikasi dan kontrol perangkat. Prototipe yang dikembangkan terdiri dari dua mode kontrol: kontrol manual untuk menyalakan dan mematikan lampu melalui antarmuka berbasis web, serta kontrol otomatis untuk pemantauan lingkungan menggunakan berbagai sensor. Hasil pengujian menunjukkan bahwa sistem ini mampu mengelola perangkat IoT secara real-time dengan efektif. Pengujian kontrol manual berhasil mengaktifkan dan menonaktifkan lampu secara jarak jauh. Pengujian kontrol otomatis untuk pemantauan tanaman mencatat suhu lingkungan stabil antara 44–46°C, kelembaban tanah pada 27%, variasi jarak sensor ultrasonik antara 2–15 cm, serta fluktuasi intensitas cahaya antara 29–120 Cd. Hasil ini membuktikan bahwa sistem dapat merespons perubahan lingkungan secara dinamis, seperti mengaktifkan pompa air saat kelembaban tanah rendah atau menyesuaikan pencahayaan berdasarkan data real-time. Dengan menerapkan arsitektur RESTful API dan komunikasi berbasis JSON, sistem ini menawarkan skalabilitas tinggi dan fleksibilitas dalam pengembangan jaringan IoT. Penelitian ini menyimpulkan bahwa sistem pengendalian dan pemantauan IoT secara terpusat meningkatkan efisiensi, fleksibilitas, serta kemudahan dalam pengelolaan perangkat, sehingga dapat diterapkan dalam berbagai bidang seperti smart home, pertanian cerdas, dan otomatisasi industri.
Multihop Data Transmission Using LoRa Technology Manalu, Istas Pratomo; Sigiro, Marojahan Timbul Mula; Simatupang, Frengki; Manik, Andreas A. P.; Sitohang, Necia G. A.; Pardede, Goldi
Piston: Journal of Technical Engineering Vol. 8 No. 2 (2025)
Publisher : Program Studi Teknik Mesin Universitas Pamulang

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32493/pjte.v8i2.48926

Abstract

The use of Long Range (LoRa) technology in Internet of Things (IoT) networks has grown rapidly to support applications that require wide coverage with low energy consumption. However, physical obstacles and indoor use often cause significant signal attenuation, reducing range and increasing energy consumption. To overcome these limitations, this study implements multihop communication using LoRa repeaters to extend network coverage. In this study, DHT11 sensors are used to measure air temperature and humidity, with data transmitted through a multihop scheme consisting of three LoRa devices: transmitter, repeater, and receiver. The purpose of this study is to analyze the performance of LoRa communication in a multihop scheme, focusing on measuring key parameters such as Received Signal Strength Indicator (RSSI), Signal-to-Noise Ratio (SNR), and Packet Loss. Point-to-point experiments showed that the RSSI was in the range of -103 dBm to -105.5 dBm, while in multihop, the recorded RSSI ranged from -102 dBm to -105 dBm. Meanwhile, the SNR in point-to-point ranged from -2 dB to -22 dB, and in multihop, the SNR value varied from -1.00 dB to -14.50 dB. At a distance of 1.5 kilometers, the point-to-point method suffers from a high packet loss of 65%, with only 23 out of 67 packets received, indicating inadequate performance. In contrast, the multihop method successfully reduces the packet loss to only 0.8%, with 33 out of 36 packets received, indicating improved data transmission quality and reliability. Tests show that the use of LoRa repeaters in multihop networks can significantly extend communication range and improve energy efficiency, with successful data delivery and performance that meets expectations. This research makes an important contribution to understanding the implementation of multihop LoRa networks, especially in the context of IoT applications that require wide coverage in congested environments.
Design of Handoff Communication Sequence Architecture in LoRa Networks Manalu, Istas Pratomo; Simatupang, Frengki; Sinambela, Eka Stephani; Sigiro, Marojohan Mula Timbul; Wowiling, Gerry Italiano; Silalahi, Sari Muthia
Piston: Journal of Technical Engineering Vol. 9 No. 1 (2025)
Publisher : Program Studi Teknik Mesin Universitas Pamulang

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32493/pjte.v9i1.51408

Abstract

Technological advances have driven the development of IoT-based object tracking systems, where LoRa is an ideal wireless technology due to its long range and low power consumption. Challenges in implementing LoRaWAN, particularly its role in the handoff process between gateways that can disrupt communication, can be overcome by developing a more efficient handoff method. For this reason, this study presents the design of Handoff communication for the LoRa Network. We use two gateways and one transmitter node. The gateway node consists of a LoRa module and an ESP32, while the Transmitter consists of a LoRa module, an Arduino Nano, and a GPS sensor. The RSSI parameter is a determining factor in transferring connectivity paths from GW A or GW B, as it provides an RSSI threshold value of -100 dBm. We successfully designed handoff communication at each Node and conducted a mini-test. The test results show that LoRa can implement handoff techniques at a distance of 0-500 meters. This indicates that the node is in closer range to GW A. The RSSI value of GW1 is in the range of -52 dBm to -98 dBm, while the RSSI of GW2 is in a much weaker range, which is around -120 dBm to -100 dBm. As the distance increases, the RSSI value of GW1 shows a significant decrease, while the RSSI of GW B actually increases. At a distance of approximately 250 meters, there is an intersection point between the RSSI values of the two gateways, marking the optimal handoff point. Thus, this system is able to select the best gateway, provide redundancy, check gateway availability before handoff, and handle handoff failures, thereby improving the efficiency and effectiveness of data delivery.
A Web-Based Navigation Control System for Lake Toba Cleaning Using NodeMCU ESP8266 and Pulse Width Modulation (PWM) Simatupang, Frengki; Manalu, Istas Pratomo; Siagian, Pandapotan; Sigiro, Marojahan Mula Timbul; Wowiling, Gerry Italiano; Sinambela, Eka Stephani; Saragih, Ripandy; Yosheva, Grace; Silaen, Romaito
Piston: Journal of Technical Engineering Vol. 9 No. 1 (2025)
Publisher : Program Studi Teknik Mesin Universitas Pamulang

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.32493/pjte.v9i1.51670

Abstract

Waste pollution in Lake Toba has become a critical environmental issue, threatening both its natural beauty and ecological sustainability. Manual waste collection methods remain limited in terms of efficiency and operational reach. This study aims to design and evaluate a web-based navigation control system for a floating surface-cleaning device utilizing the NodeMCU ESP8266 microcontroller. The system enables real-time control of direction and motor speed through a web interface, employing Pulse Width Modulation (PWM) for precise speed regulation. A prototype-based engineering approach was adopted, encompassing system design, implementation, and performance testing on land and in water environments. The experimental results indicate that the system successfully responded to all navigation commands (forward, backward, turn, pivot, and stop) with 100% accuracy under a stable local Wi-Fi network. Motor performance in water was found to be approximately 15–20% lower than on land due to fluid resistance. Battery endurance tests showed an operational time of approximately 3 hours on land and 2.1 hours in water at a 60% PWM duty cycle. Overall, the system demonstrates effective and flexible performance and holds promise for further development through the integration of sensors, camera modules, GPS-based autonomous navigation, and LoRa communication.