CHIPSET

Rancang Bangun Sistem Emergency Calling pada Pengendara Roda Dua melalui Helm

Ikhsal Gegana Nusantara — 2026-10-31

Motorcycle rider safety is a major concern due to the high number of traffic accidents. This research develops a helmet-based Emergency Calling system capable of detecting accidents and automatically sending emergency notifications. The system utilizes a piezoelectric sensor to detect impacts, a GPS module to determine the rider’s location and last recorded speed, and an ESP32 microcontroller to integrate data into a Telegram bot. Testing results indicate that the piezoelectric sensor can detect impacts with high sensitivity (ADC value ≥ 1024), the GPS module achieves 84.65% accuracy in determining speed and location, and emergency notifications can be sent to Telegram with an average response time of 0.706 seconds. This system allows families or authorities to respond promptly to accidents, increasing the chances of saving victims.

Intelligent Monitoring System for Broiler Chicken Houses Blower using Fuzzy Logic

Muhammad Fadhil Kenedy — 2026-10-31

This research presents the design and implementation of an intelligent early warning system for monitoring blower performance in Closed House broiler farms. The system aims to reduce financial losses and mitigate livestock health risks caused by poor air circulation due to blower degradation or failure. It integrates multiple sensors, including anemometers for wind speed, MQ-135 for ammonia gas, and MQ-7 for carbon monoxide. Data are processed by an ESP32 microcontroller within a distributed architecture of two monitoring boxes connected via NRF24L01 wireless modules. A fuzzy logic algorithm enables multi-variable decision-making, allowing the system to activate backup blowers, trigger alarms, and send real-time notifications through Telegram. Experimental results confirm system reliability, with anemometer accuracy ranging from 85% to 92%, buzzer intensity reaching 92–115 dB, and stable wireless communication up to 91 meters. The fuzzy logic controller successfully classified environmental conditions and blower performance, generating tiered alerts (Warning 1, Warning 2, Emergency). The findings demonstrate that the proposed system provides a proactive and intelligent solution, offering greater adaptability than conventional threshold-based methods while supporting productivity and animal welfare in poultry farming.

Single Board Computer Based Driver's Mobile Phone and Vehicle Speed Detection System with Real Time Processing

Anisha Shathya — 2026-10-31

A toll road is a type of public road that is part of the national road network in Indonesia, where users are required to pay a toll fee to access it. Toll roads are specifically designed for vehicles traveling at high speeds, so drivers are expected to always be alert and comply with the speed limits set by the toll road authorities. This research aims to design and build a mobile phone and vehicle speed detection system capable of monitoring the condition of the driver and vehicle. This system uses an NVIDIA Jetson Nano as the main processor connected to OBD-II and a Raspberry Pi Camera Module 2. This system design aims to reduce the number of accidents on toll roads by utilizing image processing using YOLO (You Only Look Once) to detect an object and pose for the driver.

Home Electricity Cost Control System Based On Internet of Things

Muhammad Rifqi Asyari — 2026-10-31

The increase in electricity consumption, especially in the household sector, is often not realized by users, resulting in uncontrolled increases in electricity bills due to the lack of transparency in real-time usage data. This study aims to design and implement an Internet of Things (IoT)-based electricity cost monitoring and control system for several electrical devices at home as a solution to overcome this problem. The method chosen based on the House of Quality (HoQ) analysis is an IoT system that uses an electrical energy sensor (then implemented as an ACS712 current sensor and a ZMPT101b voltage sensor) and an ESP32 microcontroller as a gateway. This system will measure the current and voltage values of the device in real-time, calculate power, accumulate electrical energy (kWh), and calculate cost estimates. The calculated data is sent wirelessly to the Firebase cloud database. Users can monitor this information, input the electricity price per kWh and the initial token, and control electrical devices through a website interface. In addition, the system can provide an estimate of the device's lifespan and automatically turn off the device if the token balance runs out. Test results
show that this system is able to control the electricity costs of several devices in the home effectively, so it has the potential to be a significant solution in the efficiency of household electricity use

Automated Floor Cleaning System for Hedgehog Cages using The You Only Look Once (YOLO) Algorithm for Waste Detection

Fadlan Aziz — 2026-10-31

African pygmy hedgehogs (Atelerix albiventris) have gained popularity as exotic pets; their reliance on olfactory and auditory senses makes them highly sensitive to environmental changes, particularly poor cage hygiene. Unsanitary conditions may cause odor accumulation, microbial growth, and parasitic infestations, leading to stress, reduced appetite, digestive disorders, infections, and respiratory issues. This study presents the design and evaluation of an automated cage floor cleaning system for hedgehogs based on computer vision using the You Only Look Once (YOLO) algorithm. The proposed system performs precise waste detection, autonomous cleaning, and remote monitoring and control via a mobile application. The cage consists of six PETG floor plates rotated by 180° using six MG996R servo motors, while a 12V submersible pump performs rinsing. Processing is conducted on a Raspberry Pi 4B with a PCA9685 PWM driver, integrated with Google Cloud/Firebase and a Flutter-based mobile application. Real-time video streaming is implemented using WebRTC and decoupled from the detection pipeline to maintain a stable frame rate. A YOLOv11s model was trained for object detection, achieving a precision of 0.785, recall of 0.753, and mAP@0.5 of 0.802 (mAP@0.5–0.95 of 0.491). Hedgehog detection achieved high performance (AP 0.969), while waste detection showed lower performance (AP 0.676) due to visual similarity with wooden pellets. Experimental results demonstrate smooth real-time streaming (23–27 FPS), low control latency, and reliable automated cleaning with a safety interlock mechanism that prevents cleaning when a hedgehog is detected. The system is feasible for automated cage maintenance, with future improvements focusing on dataset enhancement and system optimization.

Identification and Prediction System of Tomato Leaf Diseases Based on YOLO and Random Forest

Rafki Harzikra — 2026-10-31

Tomato crop failure in Indonesia caused by leaf disease can reach 78%, and manual visual monitoring by farmers is often delayed and subjective. This study designs and evaluates a portable, IoT-based diagnostic and prediction system for tomato leaf disease that can operate independently in the field. The system integrates a Raspberry Pi 4 single-board computer with a Raspberry Pi Camera V2, a DHT22 air temperature/humidity sensor, and a capacitive soil-moisture sensor. Two computational models run on-device: a YOLOv5 convolutional neural network that classifies leaf images into healthy, early blight, late blight, and septoria leaf spot; and a Random Forest classifier that predicts disease potential from air temperature, air humidity, and soil moisture. Results are transmitted via a 4G module to a Firebase cloud database and displayed on a real-time dashboard. Functional and field tests validated camera resolution, sensor accuracy, model accuracy, processing latency, and transmission delay. The YOLOv5-small model classified leaf images with 85.4% accuracy, while the Random Forest model predicted disease potential with 97% accuracy. On-device dual-model inference averaged 7.26 seconds and cloud transmission delay averaged 8.8 seconds, both within the 10-second design target. Field validation on a working tomato farm produced an average diagnostic correctness of 85.53%. The dual-model, off-grid-capable IoT system is technically feasible for real-time, field-deployable tomato leaf disease diagnosis and can help farmers act earlier on emerging infections, reducing potential crop loss.

Sistem Pendeteksi Penyakit dan Pemantauan Kondisi Tanah serta Rekomendasi Pengendaliannya pada Tanaman Cabai

Alhafiz Zhurrahman Mirza Utama — 2026-10-31

The cultivation of chili plants faces challenges in early disease detection and accurate soil condition monitoring, which can potentially reduce crop yields and increase the risk of plant damage. This research aims to develop a disease detection and soil condition monitoring system based on Raspberry Pi, utilizing the You Only Look Once (YOLO) version 5 algorithm to detect diseases in chili plants from images captured by a camera. The system uses Raspberry Pi 4B as the data processing center, with a camera for capturing plant images and pH and soil moisture sensors to monitor soil conditions. The data obtained is used to provide better handling recommendations for farmers. Test results show that this system can detect plant diseases with 87.5% accuracy and an F1-score of 93.3% at a distance of 5 to 25 cm, although the accuracy decreases at distances greater than 25 cm. Furthermore, soil condition measurements provide important information about soil pH and moisture, helping farmers make more informed decisions in maintaining plant health. All disease detection and soil condition information is displayed on an LCD screen connected to the Raspberry Pi, making the system user-friendly for farmers. Therefore, this system is expected to be an effective solution for increasing crop yields and reducing losses due to diseases in chili plants, and it can be more broadly applied in the agricultural sector.

Computer Vision-Based PPE Compliance Detection in a Pharmaceutical Laboratory

Abil Rahman — 2026-10-31

The use of Personal Protective Equipment (PPE) is important for maintaining practitioner safety in pharmaceutical laboratories; however, manual inspection may lead to oversight. This study aims to develop a system capable of automatically detecting the use of masks, laboratory coats, and gloves and providing detection results. The system was developed using a computer vision approach with the YOLOv8s model implemented on the NVIDIA Jetson Orin NX and trained using five classes: mask, labcoat, glove, no_glove, and no_labcoat. The detection results were used to determine PPE completeness, provide notifications through an LCD and speaker, and record violations through the Firebase Realtime Database for subsequent display on a Progressive Web Application (PWA). Evaluation results show that the model achieved 89% accuracy and 96.5% mAP@0.5, while testing on 40 samples resulted in a 95% detection success rate. The system also operated continuously for 8 hours without lag or crashes, with an FPS of 14–16 and power consumption of 5.7–9.2 W. These results indicate that the developed system can perform real-time PPE detection and monitoring with stable performance and can support PPE compliance supervision in a pharmaceutical laboratory.

Design and Development of a Helmet Drying and Sterilization System Using the Fuzzy Logic Method

Prima Gunawan — 2026-10-31

This study aims to design and develop an IoT-based helmet drying and sterilization system utilizing the Mamdani fuzzy logic method to maintain helmet hygiene, reduce humidity, and inhibit the growth of harmful microorganisms. The system integrates a DHT22 sensor to measure temperature and humidity, an ESP32 microcontroller for data processing, a fan heater for drying, and a UV-C lamp for sterilization. To handle dynamic environmental conditions, the Mamdani fuzzy logic algorithm adaptively controls the actuator outputs based on sensor inputs. Additionally, the system connects to the Blynk IoT platform via Wi-Fi for real-time remote monitoring through a smartphone application. Testing demonstrated that the DHT22 sensor achieved high accuracy, recording 99.71% for temperature and 97.17% for humidity. The ESP32 efficiently processed the fuzzy logic algorithm with an average latency of 556 µs, while maintaining stable IoT communication up to a 40-meter radius. The fan heater successfully reached the target drying temperature of 45°C, and the UV-C lamp effectively operated within the germicidal spectrum to ensure proper sterilization. Overall, the system dynamically adjusted to varying conditions and successfully completed the drying process in approximately 1 hour and 3 minutes. In conclusion, the proposed system successfully automates helmet maintenance, proving highly effective in drying and sterilizing helmets. The integration of fuzzy logic and IoT technology ensures adaptive performance and reliable monitoring, thereby enhancing user health, safety, and comfort.

Sistem Monitoring Fokus Belajar Anak Usia 5-7 Tahun menggunakan Pose Detection berbasis Mediapipe

Ani Febrian Hasanah — 2026-10-31

Penelitian ini bertujuan mengembangkan dan mengevaluasi subsistem deteksi postur tubuh berbasis computer vision untuk memantau tingkat fokus belajar anak usia 5-7 tahun. Sistem menggunakan MediaPipe Pose untuk mendeteksi 33 landmark tubuh secara real-time, dengan memanfaatkan rata-rata koordinat Y kedua bahu dan ambang batas 0,25 untuk mengklasifikasikan postur duduk dan tidak duduk. Sistem diimplementasikan pada Raspberry Pi 5 dengan webcam Logitech C270, menggunakan Finite State Machine (FSM), SQLite untuk penyimpanan data sementara, dan Firebase untuk visualisasi pada dasbor web. Pengujian meliputi pengaruh jarak kamera, intensitas cahaya, akurasi deteksi, dan kinerja sistem. Hasil pengujian terhadap 47 sampel menunjukkan akurasi sebesar 87,23%, dengan 41 sampel terdeteksi benar dan 6 sampel tidak duduk salah diklasifikasikan sebagai duduk. Landmark bahu tetap dapat terdeteksi pada intensitas cahaya 100, 200, dan 314 lux, sedangkan peningkatan jarak kamera memengaruhi keandalan deteksi. Sistem terintegrasi menghasilkan rata-rata frame rate sebesar 10,91 FPS, latensi 90,75 ms, penggunaan memori 2,03% dari RAM 16 GB, utilisasi CPU sebesar 27%, dan suhu maksimum 69,7°C. Hasil penelitian menunjukkan bahwa subsistem yang dikembangkan layak digunakan sebagai parameter awal dalam pemantauan fokus belajar anak pada perangkat embedded, meskipun akurasi deteksi masih dipengaruhi oleh kalibrasi jarak kamera dan penyesuaian tinggi tripod terhadap tinggi badan anak.

Sitting Pressure Distribution Monitoring System Using K-Nearest Neighbor and Rule-Based Logic

Athaya Thalita Jumeyza — 2026-10-31

An embedded sitting-pressure monitoring system is developed to identify asymmetric pressure distribution and detect prolonged static sitting. Four FSR402 sensors connected to a 12-bit ADC on an ESP32 measure seat
pressure, using Left_percent, Right_percent, and P_ratio as classification features. A K-Nearest Neighbor (KNN) classifier (K = 5) is implemented directly on the ESP32 to classify Balanced, Lean Left, and Lean Right postures. Classification results are combined with duration-based rules to determine NORMAL, WARNING, and HIGH RISK conditions, supported by vibration feedback, Firebase Realtime Database, and a Flutter application. The results show an average relative pressure measurement error of 1.05% (98.95% accuracy), with pressure shifts of 17.32–18.29%. KNN achieves 98.89% mean cross-validation accuracy and 97.78% test accuracy, with a macro F1-score of 0.9776, while end-to-end posture detection achieves 90% accuracy. Haptic feedback delays range from 0.784 to 1.127 seconds. These findings demonstrate the feasibility of implementing KNN-based posture classification directly on an ESP32 and integrating it with duration-based rules for real-time monitoring of asymmetric and prolonged static sitting.