Green Intelligent Systems and Applications

A Systematic Literature Review of YOLO and IoT Applications in Smart Waste Management

2025-08-28T00:56:31+00:00

The increase in urbanization and global population expansion resulted in increased garbage production, causing considerable environmental and public health issues that exceeded traditional waste management approaches. To tackle these challenges, automated waste detection and analysis integrated computer vision, especially deep learning, with the Internet of Things (IoT) in intelligent waste management applications. This comprehensive literature review investigated a wide range of You Only Look Once (YOLO) applications in IoT-based waste detection and management, demonstrating its efficacy in addressing global waste issues. Employing specific keywords and Boolean operators, the review followed a rigorous methodology to explore reputable electronic databases for peer-reviewed articles published from 2019 to 2025. The primary findings indicated that different iterations of YOLO (v3 to v12) were integrated with diverse IoT devices and computing setups, including edge and centralized systems. These integrations facilitated four crucial applications: hazardous waste management, monitoring of smart bins, classification of waste types, and detection of litter in public spaces. This integration enhanced sustainability through improved waste management practices, increased efficiency in waste processes, and reduced manual labor requirements. Challenges included precise waste identification in complex scenarios, adaptation to fluctuating environmental conditions, and ensuring dependable, low-power operation of IoT devices. To sum up, the integration of YOLO and IoT established a robust basis for intelligent waste management, transforming reactive approaches into proactive strategies. Moving forward, research should prioritize enhancing the integration and power management of IoT sensors, optimizing edge deployment, and developing more resilient YOLO models.

Maritime Cybersecurity Readiness and Training Standards in Indonesia

2025-12-25T01:29:22+00:00

The rapid digitalization of maritime operations through IoT-enabled navigation systems and cyber-physical ship infrastructures increased Indonesia’s exposure to cybersecurity risks. Strengthening cybersecurity competence within Maritime Education and Training (MET) institutions was therefore essential to ensure navigational safety, operational reliability, and national maritime resilience. This study assessed cybersecurity readiness, training standards, instructor competence, and facility availability in Indonesian MET institutions with reference to international frameworks, including IMO MSC-FAL.1/Circ.3, BIMCO guidelines, and ISO/IEC 27001. A descriptive quantitative approach was employed using structured questionnaires to evaluate organizational readiness, curriculum implementation, instructor qualifications, and supporting facilities. Data were analyzed using percentage distributions to identify institutional conditions and gaps relative to global requirements. The results indicated that cybersecurity training in most MET institutions remained largely theoretical, with limited practical exposure. Nearly 80% of respondents reported having no prior cybersecurity training, while hands-on facilities such as cyber laboratories and simulation environments were largely unavailable. Instructor expertise and standardized cybersecurity modules aligned with international guidelines were insufficient to adequately address threats to AIS, GPS, ECDIS, and integrated IT–OT systems. These findings revealed a significant gap between existing training practices and the competencies required for secure digital maritime operations. The study concluded that standardized, practice-oriented cybersecurity training was urgently needed, supported by instructor upskilling, curriculum alignment with international standards, and the development of shared training facilities. Strengthening these aspects was critical to improving national maritime cyber readiness and supporting resilient intelligent maritime systems.

Temporal Analysis of Land Subsidence in DKI Jakarta Using the Long Short-Term Memory (LSTM) Model

2025-12-25T01:29:25+00:00

This research investigated temporal patterns of land subsidence in DKI Jakarta by applying a Long Short-Term Memory (LSTM) model to deformation measurements derived from Persistent Scatterer Interferometric Synthetic Aperture Radar (PS-InSAR) observations acquired between 2017 and 2021. Because the original PS-InSAR time series was characterized by uneven acquisition intervals, the deformation records were first resampled into a uniform 11-day sequence to obtain a consistent temporal structure for modeling. Preprocessing steps, comprising outlier handling, temporal resampling, and feature normalization, were performed to ensure that the model could capture deformation behavior reliably. The LSTM architecture employed three stacked recurrent layers and was trained using the Adam optimizer with Smooth L1 Loss and an early-stopping strategy. Model evaluation demonstrated excellent agreement between predicted and observed deformation, yielding R² = 1.000, MSE = 0.104, RMSE = 0.322 mm, and KGE = 0.998. Compared with a previously developed Random Forest model (R² = 0.9995, RMSE = 0.3314 mm), the LSTM approach exhibited more stable temporal learning and was better suited for long-horizon deformation forecasting. Spatial analysis revealed that northern Jakarta, particularly Cengkareng, Tanjung Priok, and Pantai Indah Kapuk, continued to experience the greatest cumulative subsidence (> −30 mm), whereas areas in the south, such as Jagakarsa and Kebayoran Baru, showed minimal deformation (< −5 mm), aligning with known geological and anthropogenic conditions. Overall, integrating PS-InSAR time series with an LSTM framework provided a more coherent and temporally consistent method for characterizing subsidence behavior in Jakarta. The outcomes of this study offered a scientific basis for developing intelligent monitoring tools to support mitigation efforts and sustainable urban planning in regions affected by land subsidence.

Design and Build a Push-Pull Inverter for Room Lighting

2025-08-28T00:56:30+00:00

This study addressed the issue of harmonic distortion in solar power systems that required inverters to convert DC voltage to AC for indoor lighting applications. The objective was to design and evaluate a push-pull inverter incorporating pulse width modulation (PWM) to reduce harmonics and ensure a stable voltage output. A push-pull topology was selected because of its relatively simple design and ability to step up DC voltage using a transformer, making it suitable for low- to medium-power applications. The inverter employed two metal–oxide–semiconductor field-effect transistor (MOSFET) switching devices operated alternately to generate AC waves at the output. The core of the design was a 50 Hz pulse generator producing a 5 V pulse signal with a small current, which was then amplified using a current amplifier before being supplied to the transformer. The transformer functioned to induce the electromagnetic field from the pulse source and release it at a higher voltage of 220 V. Experimental testing was performed using 2.3 W, 5 W, and 8 W LED lights. A minor modification to the gate resistor improved system performance, resulting in stable transformer output voltages at 5 W and 8 W loads. These results demonstrated that the PWM-controlled push-pull inverter successfully reduced harmonics and maintained voltage stability under higher loads, making it effective for indoor LED lighting powered by solar energy. Future studies could aim to enhance efficiency at lower loads, minimize switching losses, and implement more advanced PWM techniques to achieve performance levels comparable to pure sine wave inverters.

Implementation of the You Look Only Once (YOLOv11) Algorithm to Detect the Ripeness of Golden Melons

2025-12-25T01:29:21+00:00

Melon is a horticultural commodity with high economic value, and characteristics such as sweetness, aroma, texture, and phytonutrient content significantly influenced consumer preference. Conventional methods for determining melon ripeness were time-consuming, required considerable expertise, and were often prone to subjective errors, ultimately slowing the production and distribution process. This study aimed to detect the ripeness level of golden melon fruit non-destructively using the YOLOv11 algorithm, focusing on external physical characteristics as the basis for classification. The objectives included applying transfer learning to categorize golden melon into ripe and unripe classes and evaluating model performance using precision, recall, mAP50, mAP50-95, and F1-score. The research methodology consisted of a literature review, dataset collection from previous studies, system design, implementation, and performance testing. The dataset was divided into 70% training, 20% validation, and 10% testing data, and the Adam optimizer was used during the training phase. Based on four experimental scenarios, scenario 3 produced the best and most consistent results, achieving a precision of 90.58%, a recall of 90.79%, an mAP50 of 97.31%, an mAP50-95 of 88.84%, and an F1-score of 92.97%. These findings demonstrated that scenario 3 offered optimal performance for detecting golden melon ripeness. Thus, the model was highly reliable overall.

Design an Electricity Consumption Prediction Information System Using the Monte Carlo-Based Regression Tree Method

2025-12-25T01:29:24+00:00

Electricity became an essential component in every industry and was widely used in organizations and households. Improper handling of electricity consumption resulted in unnecessary energy loss and increased costs. The objective of this study was to develop an online electricity consumption prediction information system that was efficient, reliable, and capable of rapid forecasting. The system used IoT sensor data from Universitas Widya Dharma Pontianak, and the Monte Carlo based Regression Tree (MCRT) method was employed to mitigate the unpredictability of the data. Feature selection was conducted using Monte Carlo simulation to identify the most important features, which in this case were the year, month, and day, and these were used in the regression tree model. The developed system was able to provide estimations of hourly and daily energy consumption and the associated costs based on the MCRT model. The MCRT model predicted daily energy consumption with an accuracy of 91.61%, outperforming the Monte Carlo simulation (85.39%) and the Regression Tree method (84.29%). The results demonstrated that the MCRT model was the most efficient in capturing non-linear relationships and regression patterns in the energy consumption data. The constructed system featured an easy-to-use web interface that captured real-time data inputs and visualized predicted consumption for operational use. The system was suitable for public and private sectors, as well as educational and household applications. This approach improved effectiveness in energy management and streamlined resource allocation decision-making. The study highlighted the potential of integrating the Internet of Things (IoT) with predictive analytics to provide actionable, reliable, and precise energy management and monitoring services.

Implementation of Key Performance Indicators in the Palm Oil Harvest Monitoring Information System

2025-08-28T00:56:28+00:00

Manual management of palm oil harvest data often resulted in data inconsistencies, low operational efficiency, and financial losses for plantation companies. To address these challenges, this study aimed to implement the Key Performance Indicator (KPI) method within a harvest monitoring information system at PT Perkebunan Nusantara (PTPN; Nusantara Plantation Company, Ltd) IV Regional II Unit Adolina, with the goal of enhancing data accuracy, transparency, and managerial effectiveness. A quantitative research approach was employed, utilizing data collection techniques such as observation, interviews, and document analysis. Five out of ten afdeling units were selected as research samples. The performance was assessed using three primary indicators: total harvest yield (weighted at 40%), labour productivity (35%), and monthly harvest frequency (25%). The results revealed significant variations in KPI achievement across afdeling units, with several units exceeding the established targets. Furthermore, the developed system featured an interactive visual dashboard that enabled managers to monitor performance in real time and supported data-driven decision-making. Compared to traditional monitoring tools, this system offered enhanced integration of performance metrics, automated data processing, and real-time analytics, addressing previous limitations such as delayed reporting and fragmented data sources. In conclusion, the integration of KPI into the harvest monitoring information system proved to be effective in providing objective and measurable performance evaluation. This approach offered a strategic solution for improving operational efficiency and productivity in palm oil plantation management.