Jurnal Otomasi Kontrol dan Instrumentasi

Mapping Public Emotions Regarding the 2025 UTBK Announcement in Indonesia: A Multi-Label Approach with Targeted Calibration

2025-11-16T17:51:22+07:00

This study maps public emotions in Indonesian-language tweets related to the 2025 UTBK announcement using multi-label emotion classification. The main challenges in multi-label emotion classification on social media include extreme label imbalance, distribution shift between training and application data, and weak lexical signals for specific emotions. This study aims to build a reliable emotion modeling framework for long-tail social media corpora while demonstrating generalizable post-training calibration practices. The novelty lies in the integration of four components: (1) per-label posterior calibration using Platt scaling, (2) precision-targeted per-label thresholding frozen from the development set, (3) score-quantile–based rate targeting to align predicted prevalence with domain-based rates, and (4) context-limited lexicon-aware boosting with a final clamp. The proposed pipeline is lightweight and model-agnostic. This research adopts a quantitative experimental approach by varying post-training calibration components to measure their impact on classification performance. An IndoBERTweet model is trained using BCEWithLogitsLoss on manually annotated data, then calibrated and evaluated on development and test sets. The results demonstrate balanced micro- and macro-level performance, improved detection of minority labels, and emotion mapping over 3,500 tweets with prevalence distributions consistent with Plutchik’s theory of emotions.

Design and Development of a Water Level Monitoring System on a Water Level Simulator

2026-01-02T10:43:09+07:00

Water level monitoring is generally still done manually. To improve monitoring effectiveness, a real-time water level monitoring system was designed using an aquarium simulator. This simulator represents three scenario conditions: Normal, Alert, and Danger. This system uses an ultrasonic sensor to measure water levels, a microcontroller as a data processor, and a wireless communication module to transmit data to a central server. The collected data is displayed on a local display for easy access by readers. Through testing, this system demonstrated an accuracy of 96,43% with a swift data transmission time of 0,05-0,07 seconds. The implementation of this system provides an effective automated monitoring solution, minimizing the need for manual observation, and this system has the potential to improve flood risk mitigation through more accurate and responsive early warnings.

Defect Detection System in Coffee Beans Using Roboflow-Detection Transformer (RF-DTER) Algorithm

2025-11-09T06:51:53+07:00

Coffee bean quality control is a critical stage in processing industries to meet export and consumption standards. Traditional visual manual inspection often results in inconsistency, subjectivity, and reduced production throughput. This research implements the Roboflow Detection Transformer (RF-DETR), an end-to-end transformer-based object detection architecture, to identify subtle and complex coffee bean defects. The study uses image processing and machine learning with a labeled dataset of 2,010 coffee bean images classified into five defect categories: brown, black, unripe, broken black, and partially black. The data are split into 75% training, 17% validation, and 8% testing. Performance evaluation shows RF-DETR detects and classifies all defect types effectively, achieving a mean Average Precision (mAP) of 97,6%, with 95,7% precision, 91,0% recall, and an F1 score of 93,29%. These results indicate that RF-DETR balances accurate spatial localization with reliable class prediction, minimizes false positives, and maintains strong detection sensitivity. Therefore, RF-DETR provides a solid technological basis for high-precision, real-time automated coffee bean sorting in industrial settings. For deployment, it can be integrated with production cameras and conveyor sorting actuators to deliver fast, consistent decisions. Future work may optimize augmentation, lighting calibration, and edge computing deployment to improve robustness across varied production lines in practice.

Model Reference Adaptive Control for pH Neutralization in Batik Wastewater Treatment

2025-10-25T08:30:37+07:00

The batik industry generates wastewater with elevated pH levels due to sodium hydroxide usage in dyeing processes, frequently exceeding the regulatory standard of pH 6–9. Traditional dilution methods prove economically inefficient, necessitating more adaptive control strategies. This research designs and simulates a nonlinear adaptive control system based on Model Reference Adaptive Control (MRAC) for pH neutralization using acetic acid. The Continuous Stirred Tank Reactor (CSTR) mathematical model was developed from mass balance, acid-base equilibrium, and electroneutrality principles, then simulated using Ordinary Differential Equation (ODE) functions in MATLAB. MRAC performance was compared with conventional PI controller across various initial wastewater pH conditions. Simulation results demonstrate that MRAC achieves faster convergence, reaching pH 7 in 2488.68 seconds from initial pH 9, compared to PI controller (3080.96 seconds) or uncontrolled system (10225 seconds). With settling time of 2377 seconds versus 2566.4 seconds for PI, MRAC reduces neutralizer consumption by 0.61% (9.5380 L versus 9.5969 L) while maintaining safety criteria above the lower pH bound of 6.80. Lyapunov stability analysis confirms the asymptotic stability of the adaptive controller. This study demonstrates that MRAC offers superior performance for batik wastewater pH control while reducing dependency on uneconomical dilution methods.

Design of an Electrochemical-Based Biosensor Strip for Blood Sugar and Uric Acid Testing

2026-01-10T19:40:50+07:00

Rapid and accurate monitoring of glucose and uric acid levels is essential for the early detection of metabolic disorders such as diabetes mellitus and hyperuricemia. This study aims to develop an electrochemical biosensor strip based on a conductive carbon electrode for the detection of glucose and uric acid using glucose oxidase and uricase enzymes. The strip was designed with a three-electrode configuration on an acrylic substrate, which was selected due to its lower resistance and superior coating stability compared to PVC. The signal detection system employed a transimpedance amplifier circuit using the LM358, integrated with an Arduino Uno. Measurements were performed using artificial blood samples with various glucose concentrations (0–200 mg/dL) and uric acid concentrations (0–12 mg/dL). Calibration results demonstrated a linear relationship between analyte concentration and the output current, with R² values approaching 1. After correction, the glucose strip exhibited 92–100% accuracy and 92–97% precision, while the uric acid strip achieved 76–99% accuracy and 96–97% precision. These findings indicate that the developed biosensor strip provides consistent and reasonably accurate measurements, demonstrating its potential as an economical and portable point-of-care device for monitoring glucose and uric acid.

Development of a Twin-Motor Aeropendulum with Fuzzy Logic Controller Based on MATLAB-Simulink and Arduino

2026-03-31T07:14:48+07:00

This paper presents the design and development of a twin-motor aeropendulum (half-quadcopter) system as a control systems learning platform using a fuzzy logic controller (FLC). The system is implemented using a real plant in the form of a half-quadcopter, an IMU MPU6050 sensor combined with a complementary filter for angle estimation, and an integration of Arduino and Matlab-Simulink for controller design and response visualization. Compared to conventional PID controllers, the FLC is selected due to its ability to handle system nonlinearities without requiring an accurate mathematical model. The main contribution of this work lies in the development of a low-cost educational platform that integrates a real plant, a non-contact sensor with improved durability, and real-time system response visualization. Experimental results at a 30° setpoint show a rise time of 0.5 s, a settling time of 5.8 s, a maximum overshoot of 21.3%, and a steady-state error of 0.2%, along with good disturbance rejection capability. These results demonstrate that the proposed system is effective and suitable as an interactive learning platform to enhance students’ understanding of control system concepts.

Integration of Sensor-Based IoT Automation for Feeding and Cleaning in Smart Bird Cage Systems

2026-03-13T18:08:07+07:00

Pet songbirds require consistent care, but prolonged owner absence often leads to irregular feeding and poor sanitation. Existing IoT-based bird care systems remain fragmented, focusing primarily on single aspects like feeding, while sanitation mechanisms are often mechanically complex or ignored. This study proposes an integrated IoT autonomous bird cage that uniquely combines feeding, watering, and a simplified stepper-motor-driven brush-and-spray cleaning mechanism into a single framework. Controlled by an Arduino Uno R4 WiFi and an RTC module for precise offline scheduling, the system utilizes a robust capacitive sensor for water monitoring to overcome the unreliability of standard acoustic sensors in confined micro-climates. Experimental testing demonstrated sub-centimeter sensor accuracy and a 0% failure rate for automated water refilling, outperforming the 20% failure rate observed in the ultrasonic feed sensor. Furthermore, the novel cleaning subsystem achieved a 94% quantitative waste removal efficiency. These results confirm the proposed architecture is a reliable, comprehensive solution that significantly reduces manual intervention in smart pet management.

Design and Implementation of an IoT-Based Water Quality Monitoring System Using Salinity and Temperature for Malaria Mosquito Larval Habitat Risk Identification

2026-04-09T09:38:05+07:00

Malaria remains a public health problem in Indonesia, particularly in endemic coastal regions where aquatic habitats serve as breeding sites for Anopheles larvae. This study develops a multi-point Internet of Things (IoT)-based aquatic environmental monitoring system to detect conditions that support larval development through real-time measurement of water salinity and temperature. The system employs a WQ7706D digital salinity sensor, an ESP32 microcontroller, and a low-power NRF24L01 wireless communication module. Laboratory testing indicates that the sensor achieves stable readings after a 20-second stabilization period, with salinity variation of ±0,05 ppt under steady conditions. Field implementation at two coastal water sites in Hanura Village recorded salinity ranges of 1,35–2,3 ppt and temperature ranges of 28,8–30,5°C, which potentially support larval development. The wireless communication system successfully transmitted data up to 150 m with minimal packet loss. Power consumption analysis shows a daily energy requirement of 1,794 Ah, enabling autonomous operation for 10 ± 1 days using a 12 V 20 Ah battery without recharging. The main contribution of this research is the design of an IoT-based aquatic monitoring system that integrates energy optimization, stable wireless communication, and quantitative identification of malaria larval habitat risk.

Design and Construction of Alginate Mixer Control System Based on Arduino and Nextion Display

2026-04-08T15:06:17+07:00

This study aims to design and implement an Arduino-based alginate mixer control system with Nextion Display as the user interface. This system was developed to improve the accuracy of mixing time and maintain the consistency of mixing results in the production process. The method used is a design with an experimental approach through testing several variations of stirring time and comparison with a conventional manual timer-based system. The parameters observed include time error and viscosity values of the mixing results. The test results show that the Arduino-based system has better time accuracy with smaller and more stable error values. This has an impact on more consistent viscosity results, which values obtained are in the range of 100–110 dPa·s according to process standards. In addition, the developed system also provides ease of operation and monitoring through real-time displays on the Nextion Display. Thus, this system can be a more effective alternative compared to conventional systems in the alginate mixing process.

Design and Development of an IoT Based Automated Soil Water Content and Temperature Monitoring System

2026-02-25T15:35:09+07:00

Modernization of meteorological observation systems at the Agency for Meteorology, Climatology, and Geophysics (BMKG) is necessary, particularly for soil temperature and soil water content parameters that are still commonly measured using conventional methods. This study presents the design and evaluation of an Internet of Things (IoT)-based monitoring system to improve observation efficiency. The system integrates PT100 sensors to measure soil temperature at depths of 5, 10, 20, 50, and 100 cm, and Capacitive Soil Moisture v2.0 sensors to measure volumetric soil water content at depths of 10, 20, 30, 50, and 100 cm. Data is transmitted in real time via Wi-Fi to Google Spreadsheet and a Telegram bot, enabling remote monitoring and automatic notifications, while also being stored locally on an SD card for reliability. Sensor performance was evaluated through laboratory calibration in a temperature chamber and an 11-day field deployment at the BMKG Regional II Headquarters. Results show good accuracy, with RMSE of 0.93 °C (MAE 0.66 °C) at 5 cm depth and RMSE of 0.43 °C (MAE 0.29 °C) at 10 cm depth, demonstrating reliable system performance.

Design of Adaptive Traffic Light Control System Using Fuzzy Logic, Based on Number of Vehicles, Length of Queue and Types of Vehicles

2026-03-02T15:49:08+07:00

Congestion at signalized intersections is often triggered by the use of fixed-time controllers that are unresponsive to the dynamics of traffic volume, queue length, and vehicle heterogeneity. This study aims to design an adaptive traffic light control system based on a Mamdani-type Fuzzy Logic Controller (FLC) by integrating three input variables: the number of vehicles, queue length, and vehicle type/size. The novelty of this research lies in the use of the vehicle type variable as an explicit input within the FLC model. The system was designed by determining the universe of discourse, membership functions, and a base of 27 IF-THEN rules, which were then implemented in Python using the scikit-fuzzy library through MIN-MAX inference mechanisms and Centroid of Area (COA) defuzzification. Testing was conducted on 27 input combinations representing traffic conditions ranging from low to heavy. The simulation results show that the generated green light duration ranges from 5.09 to 91.60 seconds, with a tendency to increase alongside the rise in the number of vehicles, queue length, and vehicle size. Thus, the proposed model is capable of generating green light duration decisions that are more responsive and representative of mixed traffic conditions.

Development of a Desktop Application for an Automated Storage System Based on Robotics and Computer Vision

2026-04-14T10:46:57+07:00

Conventional storage and management processes in industrial sector are highly vulnerable to risks such as data entry errors, disorganized item placement, and time inefficiency due to human intervention. This phenomenon necessitates digitalization through precise and integrated automation. This research aims to design, implement, and test a application as the central control unit. The novelty of this study is the development of an application architecture that functions as a centralized supervision. This system not only manages the database but also coordinates storage workflows in real time, facilitates two-way industrial communication with FINS protocol, and performs position validation via visual detection. The research method applied is engineering research, using a research-by-project approach. The research stages are systematically, encompassing requirements identification, architecture design, program implementation, hardware integration, and comprehensive functionality testing. The application was developed using JavaFX, integrated with an Omron PLC and a computer vision module based on the YOLOv8 algorithm. Test results demonstrate that the application can perform administrative supervision functions, synchronize data communication with the PLC, and integrate the computer vision module with optimal stability. The implementation of this system is proven to minimize human error and improve the accuracy of goods placement through visual feedback mechanisms and centralized data.

Design and Evaluation of a Low-Cost Automated Sunshine Duration Recorder with Solar Cell Based on Internet of Things

2026-01-30T08:03:30+07:00

Sunshine duration (SD) is an important meteorological parameter for climate analysis and solar energy planning. In Indonesia, SD measurement still heavily relies on manual Campbell–Stokes (CS) recorders, which are prone to subjective errors, labor-intensive, and unreliable under cloudy conditions. This study develops a low-cost IoT-based SD logger using a solar cell and an INA219 sensor to estimate direct solar radiation. An ESP32 microcontroller processes the data, supported by an RTC module and an SD card for timestamping and local storage. Calibration was performed against Copernicus Atmosphere Monitoring Service (CAMS) satellite data, applying the WMO pyrheliometric threshold of 120 W/m². Remote access is provided via a Telegram bot. A 14-day field test yielded RMSE = 104.8 W/m², MAE = 56.1 W/m², and stronger correlation against CAMS (R² = 0.640, r = 0.802) than CS (r = 0.749). The mean daily SD difference was 1.09 hours against CAMS and 2.03 hours against CS. Binary classification for sunshine detection achieved an accuracy of 88.4%, precision of 79.3%, recall of 77.2%, and F1-score of 78.2%. This prototype offers an automated, accurate, weatherproof, and remotely accessible alternative to conventional CS recorders, with strong potential to advance SD monitoring and modernize meteorological infrastructure in Indonesia.

Smart Agricultural Monitoring and Automated Irrigation System Using Internet of Things (IoT) Technology

2026-04-28T13:33:36+07:00

Efficient irrigation management and real-time environmental monitoring are crucial for improving agricultural productivity and optimizing water use, especially in small-scale and urban farming systems. This study presents the development and experimental evaluation of an Internet of Things (IoT)-based environmental monitoring and automated irrigation system for agricultural applications. The system integrates a NodeMCU ESP8266 microcontroller with capacitive soil moisture, soil pH, and DHT11 temperature sensors to monitor key environmental parameters. Data are transmitted via Wi-Fi to the Blynk cloud platform, enabling real-time monitoring through a smartphone interface and a 16×2 LCD display. Automated irrigation is implemented using a relay-controlled 5V DC water pump that operates based on predefined soil moisture thresholds. Experimental validation was conducted by comparing sensor measurements with a standard reference instrument (ETP-302). The results indicate average errors of 3.4% for soil moisture, 4.1% for soil pH, and 2.8% for temperature measurements. The irrigation system successfully activated within the dry soil threshold range (430–520 ADC) and stopped automatically when sufficient moisture was reached, reducing unnecessary water consumption. Overall, the proposed low-cost IoT-based system demonstrates reliable performance, acceptable accuracy, and effective irrigation control, supporting smart agriculture and sustainable precision water management.