SINERGI
Not a member yet
563 research outputs found
Sort by
A predictive safety and maintenance framework for railway locomotives: integrating HAZOP, FMEA, and IoT-based risk mitigation
Full text linkSafety and maintenance efficiency are critical challenges in the railway industry, particularly in the use of lifting jacks for locomotive maintenance. This study proposes a predictive maintenance framework that integrates the Hazard and Operability Study (HAZOP), Failure Mode and Effects Analysis (FMEA), and Internet of Things (IoT) technology to detect potential failures in real time. A case study was conducted at a locomotive maintenance depot in Indonesia, where several occupational accidents had been recorded due to lifting jack malfunctions. Based on HAZOP and FMEA analyses components such as stoppers and drive motors were identified as having high Risk Priority Numbers (RPN), each reaching 512, indicating significant failure risks. The proposed IoT system employs HCSR-04 and MPU6050 sensors to accurately monitor the height and inclination of the equipment. Evaluation results show that the system effectively detects anomalies with minimal data deviation and a low data loss rate during a 10-day testing period. The implementation of this system significantly reduces workplace accident risks, improves maintenance efficiency, and supports digital transformation within the industrial environment. These findings demonstrate that the integration of HAZOP, FMEA, and IoT is effective for risk mitigation and can be replicated in other railway components. Moreover, this research opens new avenues for developing AI-based predictive systems and implementing digital twins as part of future smart maintenance strategies
Development and performance evaluation of an automatic size-sorting system for catfish seeds using photodiode sensors
No full textIn catfish farming, uniform seed size is crucial for ensuring balanced growth and minimizing competition for feed. Generally, size sorting is performed manually through visual observation and net separation, which is labor-intensive, time-consuming, and often causes stress or injury to fish. To address these limitations, this study aimed to develop and evaluate a real-time, low-cost automatic sorting system for live catfish seeds. The proposed system utilizes photodiode sensors and an Arduino-based microcontroller to detect variations in fish body length by interrupting a laser beam. Four photodiodes were arranged at specific distances to classify fish seeds into four size categories (<7 cm, 7–8 cm, 9–10 cm, and 11–12 cm). After classification, the system automatically directed each seed into the corresponding container. The results showed that the prototype successfully classified and sorted catfish seeds with an overall accuracy of 67.5%. In contrast, tests with PVC pipes under controlled conditions achieved 100% accuracy. These findings highlight the novelty of integrating size detection and direct sorting for live fish seeds, a feature not previously reported in the literature. Beyond its current limitations, this system provides a methodological framework for sensor-based aquaculture automation, offering potential for further improvements in accuracy, robustness, and application to other aquaculture species
Land cover changes, built-up and vegetation density, and the Urban Heat Island (UHI) phenomenon in Pekanbaru City
Full text linkPekanbaru city has a high population growth rate and is currently experiencing rapid urbanization, which is driving urban expansion. Urban development alters land cover patterns and reduces environmental quality. The development of residential areas and infrastructure reduces vegetation, affecting Land Surface Temperature (LST) and contributing to the emergence of the Urban Heat Island (UHI) phenomenon. This study aims to analyze changes in land cover, examine the correlation between LST and the Normalized Difference Built-up Index (NDBI) and the Normalized Difference Vegetation Index (NDVI), and then investigate the UHI phenomenon in Pekanbaru City. The research method is quantitative, using data from Pekanbaru City, an administrative map, and Landsat 8 OLI/TIRS imagery, which were spatially analyzed in ARGIS and QGIS. The novelty is the use of guided classification and maximum likelihood algorithms for land cover classification, which revealed significant changes over the five years from 2018 to 2023 in Pekanbaru City. Over 5 years, land cover in the city of Pekanbaru changed, with water bodies increasing by 23%, palm areas increasing by 5%, built-up areas increasing by 34%, and vegetation increasing by 10%, while bare land decreased by 57%. There are significant changes in built-up and vegetation density. The correlation between land surface temperature and built-up density is positive; however, it is negatively correlated with vegetation density. There is an urban heat island phenomenon in Pekanbaru City, characterized by surface temperatures exceeding the UHI threshold.
Real-time Unmanned Surface Robot (USR) for river quality monitoring systemm
Full text linkA real-time Unmanned Surface Robot (USR) for river water quality monitoring system is a technology that employs a small autonomous boat outfitted with sensors and other monitoring equipment to gather and transmit data on various water quality parameters like pH, temperature and total dissolved solids sensors in rivers and other bodies of water. The USR can traverse the river, gather information or data at specific points or designated locations, as well as continuously monitor a specific stretch of river at all times. The data or information was sent in real time to a central monitoring station, where it was analyzed and used to identify potential water quality problems. Initially, the USR was designed using SolidWorks software, and its structural performance was the main focus of the investigation and examination of the design. This USR was then created and manufactured. The entire USR system could help detect and mitigate pollution and other environmental problems, as well as offer useful information for managing water resources. Next, to determine the overall performance of the USR, five experiments and autopilot accuracy tests were performed. Finally, this study also verified and validated the accuracy of water quality monitoring sensors.
Towards enhanced acoustic fan booster damage detection: a comparative study of feature-based and machine learning approaches
Full text linkMachine failure detection frequently uses non-destructive monitoring techniques such as vibration analysis. Although vibration analysis can identify machine degradation, the apparatus is often costly and necessitates specialist knowledge. Additionally, many existing methods in audio classification rely on characteristics represented as pictures or vectors, which increases computational complexity. In contrast, this research introduces a novel method that substitutes vibration data with a singular numerical feature derived from audio signals, addressing both cost and complexity issues. Our objective is to develop a rapid and precise audio-based method for detecting machine damage. The acoustic signals from the machine apparatus were classified into three categories: normal, belt damage, and combined belt and bearing defect. The data processing technique involved lowering the sample rate and segmenting the data to improve computational efficiency and classification performance. We use the Welch method and appropriate statistical techniques to analyze Power Spectral Density (PSD). The performance of seven classifier models, KNN, LDA, SVM, NB, ANN, RF, and DT, was evaluated using accuracy, precision, sensitivity, specificity, and F-score. LDA achieved the highest accuracy at 92.83%, followed by ANN (92.75%), NB (92.74%), and DT (92.34%). These models outperformed KNN (89.90%) and RF (89.40%), with SVM recording the lowest accuracy at 85.40%. LDA was highly effective, achieving the highest accuracy with a single average PSD-type feature, showcasing its robustness in machine defect diagnosis. Compared to previous methods, this approach simplifies feature extraction, reduces computational demands, and maintains high diagnostic performance, providing notable benefits in terms of effectiveness and precision.
Towards low-carbon ammonia: simulation and economic evaluation of blue ammonia with carbon utilization
No full textThe proposed blue ammonia production considers technical, environmental, and economic aspects. The design of the blue ammonia using CCUS (Carbon Capture, Utilization, and Storage) technology in this study contributes to reducing carbon emissions and providing a more environmentally friendly ammonia supply in East Java, Indonesia, due to the availability of raw materials and geological storage locations for CO2 storage. Technically, the blue ammonia production was simulated with Aspen Hysys V.14.0. uses the Kellogg process, where the ammonia converter operates at a temperature of 437.60 °C and a pressure of 141.9 bar. From the environmental aspect, as much as 68.34 tons/h of ammonia produced produces CO2 71.36 tons/h, which is a total emission of 1.06 tons CO2/ tons NH3. In this study, CO2 delivery with a pipe length of 85 km (ID:539.8mm; OD: 558.7mm) was simulated using default parameters in Aspen Hysys V.14.0. In economic calculations from APEA (Aspen Process Economic Analyzer), the manufacture of blue ammonia designed in this study is very large, with a TAC (Total Annual Cost) of 93.28x108/ tons NH3. This study demonstrates the integration of CCUS technology into ammonia production, resulting in a reduction of CO₂ emissions by 1.06 tons CO₂ per ton of ammonia produced. The proposed system provides a practical approach for improving the environmental sustainability of industrial chemical processes
Risk-based predictive maintenance of medium voltage network switching equipment using analytical hierarchy process as an analytical tool
Full text linkPredictive maintenance has become crucial for enhancing the reliability and efficiency of electrical systems, especially for Medium Voltage Network (MVN) switching equipment, which plays a key role in electricity distribution. This study aimed to develop a risk-based predictive maintenance model for MVN switching equipment using the Analytical Hierarchy Process (AHP) for maintenance prioritization, along with Z-score and Monte Carlo simulation methods to evaluate risk likelihood and impact. The Z-score method assessed the probability of risks occurring, revealing a probability exceeding 90% for specific equipment, such as UP2D.2025.C4, at 93.12%. The Monte Carlo simulation assessed the potential impact of these risks, showing severe consequences for various types of equipment. For example, UP2D.2025.C1 had a mean of 28.51 and a standard deviation of 3.50, while UP2D.2025.C8 had a standard deviation of 33.17, with an impact of over 61.53%. AHP was used to assign priority weights to components based on criteria such as equipment age, operational condition, and failure history. The analysis indicated that the Lightning Arrester had the highest maintenance priority at 26.04%, followed by the Fuse Cutout at 20.62% and the Pole-Mounted Circuit Breaker at 11.15%. This research was expected to significantly contribute to the development of more efficient and effective maintenance strategies for electrical systems, particularly in the electricity distribution sector
Stress-Strain Behavior and Residual Strength of Petobo Sand with Variable Fines Content after the 2018 Palu Liquefaction
No full textLiquefaction was one of the main causes of severe ground deformation during the 2018 Palu earthquake, particularly in the Petobo area where large-scale flow liquefaction occurred. Despite extensive field investigations, no laboratory-based residual shear strength data for Petobo soil have been available to explain the exceptional mobility of the flowslide. This study investigates the stress–strain response and residual strength of Petobo silty sand containing different proportions of fines through monotonic Consolidated Undrained (CU) triaxial tests. Reconstituted specimens were prepared using the moist tamping method with fines contents of approximately 9% and 26.4%, and tested under three levels of initial mean effective stress.The higher-fines specimen generated excess pore water pressure more rapidly and showed stronger contractive behavior at small strains. However, at larger strains, the 26.4% fines specimen mobilized greater residual strength, expressed as the ratio of residual deviator stress to initial mean effective stress (0.79-0.94), compared with the 9% fines specimen (0.85-0.89). These results indicate that fines increase contractive tendency during initiation but enhance resistance against large post-liquefaction deformation. The influence of fines on post-liquefaction behavior is therefore nonlinear and dependent on deformation stage and initial density. This study provides the first laboratory-based residual strength data of Petobo silty sand within the Critical State Soil Mechanics framework, clarifying its post-liquefaction resistance characteristics. The findings improve understanding of flow behavior in silty sands, offer mechanistic insight into the 2018 Petobo failure, and supply essential parameters for calibrating constitutive models and supporting liquefaction hazard mitigation in similar alluvial deposits in Central Sulawes
An Examination of the Fe₃O₄ nanomaterial impact in conjunction with Magnetorheological Elastomer material
No full textMagnetorheological elastomer (MRE) is an advanced material class that can be used for vibration damping. This material possesses the ability to reduce vibration disturbances through adjustment of its mechanical properties in response to a magnetic field applied from an external source. The objective of this study is to ascertain the effect of incorporating Magnetite (Fe₃O₄) nanomaterials into MRE. It is expected that this new material will be more sensitive to magnetic fields in damping vibrations, which would be a significant improvement. MRE is composed of carbonyl iron powder (CIP), silicone oil, and silicone rubber, with weight proportions of 30%, 5%, and 65%, correspondingly. The addition of magnetite nanomaterials to MRE occurred at weight ratios of 0.5%, 1%, 1.5%, and 2%. Observations of this new material included elemental composition analysis and viscoelastic testing of various mixture formulations in the laboratory. From this research, it can be concluded that an MRE containing Fe₃O₄ nanomaterials has been created. For the attenuation of vibrations within the 1–100 Hz frequency range. MRE-2 (MRE with 0.5% Fe₃O₄ added) is the best choice as the primary material, as it exhibited the highest tan delta value and strong damping performance at an intermediate frequency. MRE-1 sample was used as a base material mixture without added Magnetite also an excellent choice, offering high stiffness and good damping capability at low frequencies. It is shown by the results of this experiment that the effectiveness of MRE in reducing vibration can be increased by adding Magnetite, even in the limited mid-frequency range of 0 to 100 Hz
Dynamic modeling of lithium-ion battery degradation using data-driven and physics-informed method
Full text linkAccurate real‑time prediction of lithium‑ion battery (LIB) capacity degradation is essential for embedded battery‑management systems. Equivalent circuit models (ECMs) run quickly but lose accuracy over time, whereas purely data-driven networks achieve high precision at a high computational cost. This study introduces a physics‑informed neural network (PINN) that embeds the differential equations of a first‑order Thevenin ECM directly into the loss function. Using only terminal voltage and current as inputs, the network simultaneously estimates internal resistance, polarization resistance, polarization capacitance, open‑circuit voltage, and capacity loss. The model was trained and evaluated over 300 charge–discharge cycles of a 18650 lithium-ferrous phosphate (LFP) cell. The resulting capacity degradation estimation achieved a root mean squared error (RMSE) of 0.012 and a mean absolute percentage error (MAPE) of 0.974 %, surpassing a neural ordinary differential equation baseline with RMSE of 0.215. The trained network contains 261 parameters, requires 0.6 ms per sample for inference, and consumes 49 MB of memory. This computation cost is far lower than that of a long short‑term memory (LSTM) benchmark with comparable accuracy. In addition, the proposed model maintains its accuracy under limited dataset conditions. With a fourfold smaller training set, the PINN maintained an RMSE of 0.023, whereas the LSTM error increased to 0.72. The results demonstrate that lightweight neural networks guided by physics-based constraints can provide reliable, real-time health estimation on resource‑limited hardware