Engineering Science Letter
Not a member yet
68 research outputs found
Sort by
Optimization of Malaria Cell Image Classification Using Pretrained Resnet50 Architecture with Data Augmentation and Fine-Tuning
Full text linkMalaria remains a significant health concern, particularly in tropical regions such as Indonesia, where timely and accurate diagnosis is crucial for reducing transmission and mortality. Conventional diagnosis through microscopic examination is labor-intensive, time-consuming, and highly dependent on expert availability. This study proposes an automated malaria cell image classification model using a deep learning approach based on the pretrained ResNet50 architecture. The research framework adopts the SEMMA (Sample, Explore, Modify, Model, Assess) methodology to structure the development workflow. A total of 27,558 labeled blood cell images comprising two balanced classes, Parasitized and Uninfected, were used for training and evaluation. Two model configurations were tested: a baseline model without data augmentation or fine-tuning, and an optimized model that integrates both. Augmentation techniques such as rotation, flipping, shearing, zoom, and brightness adjustment were applied to increase data diversity, while fine-tuning involved unfreezing the last 20 layers of ResNet50 to adapt pretrained features to the malaria domain. Performance was evaluated using accuracy, precision, recall, F1-score, loss, and AUC-ROC. The optimized model achieved 97.63% accuracy, 0.996 AUC-ROC, and 0.2472 loss, outperforming the baseline accuracy of 92.84%. An ablation study analyzed the individual contributions of augmentation and fine-tuning, showing that both techniques play complementary roles, with fine-tuning having the greater impact. A McNemar test confirmed that the improvements were statistically significant (p < 0.05). These findings demonstrate that the optimized ResNet50 model is effective for malaria detection and holds promise for integration into real-time diagnostic systems in resource-constrained environments
"Where Am I": Characterization of Ultra-Wide Band Real Time Location System in Line of Sight
Full text linkAbsolute position measurement is a necessity in an Autonomous Indoor Mobile Robot (AiMR) built as to compensate the systematic and non-systematic errors caused by the odometry. This paper is set out to characterize the range bias of DecaWave Trek 1000 in Line of Sight (LOS). DecaWave Trek 1000 is a commercially available Real Time Location System (RTLS) based on Ultra-Wide Band (UWB) technology. The main advantage of using such a system is it provide quick solution with best-in-class accuracy of up to +10 cm. The system also compliant with IEEE802.15.4-2011 standards and implemented based on CMOS technology. To validate its performance, four distinct environment scenarios were selected based on their spatial occupancies to represent both favourable and non-favourable operating conditions. A total of 120,000 positioning data were measured and recorded throughout this study. Firstly, the collected data were compared to the ground truth and subjected to statistical analysis to determine whether environment factors significantly influenced ranging errors. Then the range error the results confirm that while the system performs within its specified accuracy under favourable conditions, the non-favourable condition on the other hand shows a drop of 20% in ranging accuracy. Throughout this extensive characterization, a systematic error model and noise estimation were formulated, providing a critical foundation for integrating UWB RTLS measurements into advanced probabilistic localization framework. Looking forward, the developed characterization function, offers strong potential for fusion with filtering algorithm such as Bayesian filters to enhance indoor navigation reliability for an AiMR application
Analysis of Factors Affecting ERP Implementation Performance Using the Analytical Hierarchy Process (AHP): A Case Study at PT Multi Makmur Investama (Multives) Tangerang
Full text linkThe Indonesian packaging industry faces mounting pressure to align enterprise systems with evolving strategic and operational demands. This study investigates the performance of a mature Enterprise Resource Planning (ERP) system at PT Multi Makmur Investama (Multives), leveraging the Critical Success Factors (CSF) framework and Analytic Hierarchy Process (AHP) to identify and prioritize key determinants of ERP success. Ten CSFs were evaluated through expert pairwise comparisons and cross-validated via a perception survey of 51 ERP users. Results reveal ERP system quality (18.16%), top management support (14.19%), and user training (10.79%) as the most influential drivers. However, notable discrepancies emerged between expert prioritization and user satisfaction particularly in vendor support and training indicating underlying misalignments in long-term ERP usage. The study contributes a dual-layered evaluation model combining structured expert judgment and user-based validation, offering actionable insights for ERP optimization in emerging market contexts and extending theoretical discourse on ERP maturity evaluation
Development and Evaluation of a Laminated Bamboo Frame for Electric Scooters: A Preliminary Prototype
Full text linkThis study presents the design, fabrication, and evaluation of a hybrid electric scooter frame constructed from laminated Gombong bamboo and aluminium. The objective was to address environmental concerns in personal mobility by reducing reliance on high-carbon materials and improving battery safety. A CAD-based design was executed in Autodesk Inventor using bamboo beams for primary structure and aluminium joints for mechanical stability. Laminated bamboo was processed through drying, chemical treatment, and gluing. Mechanical properties were evaluated via tensile and compressive tests, achieving a tensile strength of 289 MPa. A 72V lithium-ion battery pack composed of 18650 cells was assembled with BMS integration. The resulting prototype exhibited a total frame weight of 4.8 kg and a 30 km range per charge. The integration of laminated bamboo with aluminum joints provides preliminary evidence of feasibility and highlights bamboo’s potential as a renewable material for lightweight mobility applications. These results provide an initial prototype-level demonstration of laminated bamboo as a viable material for sustainable transport applications
Works Rescheduling in Container Terminals
Full text linkContainer transportation plays a vital role in freight movement, with container yards serving as critical hubs for operations. This paper focuses on the rescheduling of handling equipment in response to unexpected crane failures, which disrupt planned tasks. To address this challenge, a heuristic algorithm is proposed, incorporating the Longest Processing Time and List Scheduling procedures. Two rescheduling strategies are evaluated through numerical experiments. The results indicate that the second strategy outperforms the first in most cases by improving efficiency, albeit with more changes to the initial schedule. The numerical examples demonstrate the effectiveness of the algorithm, achieving feasible rescheduling solutions within less than one second for all tested cases. These findings highlight the algorithm’s efficiency and practicality in real-world container yard operations
Determination of the Thermophysical Properties of Tuna Using the Heat Pulse Method
Full text linkAlthough the tuna industry plays an important role in supporting the economy and local communities, it remains underdeveloped and faces several challenges, according to reports from the Ministry of Industry and Trade and the Vietnam Tuna Association. This study focuses on examining the heat-related physical properties of tuna—specifically, how well it conducts heat (thermal conductivity), how quickly heat spreads through it (thermal diffusivity), and how much heat it can store (specific heat capacity). These properties were measured using the heat pulse method within a temperature range of -15°C to 20°C. The results showed that thermal conductivity ranged from 0.394 to 1.103 W/m·K, thermal diffusivity from 1.11×10⁻⁸ to 2.60×10⁻⁷ m²/s, and specific heat capacity from 3,377.21 to 52,948.54 J/kg·K
BERT Model Implementation for Dynamic Sentiment Analysis of Pertamina on Social Media X
Full text linkThis study aims to investigate the dynamics of public sentiment on platform X in response to the Pertamina corruption scandal, exploring how trust and perception shifted before and after the incident. Utilizing BERT-based sentiment classification model trained on real-world social media posts, the model achieved a validation loss of 0.5078 and an F1-score of 82.12%, demonstrating strong predictive performance for large-scale sentiment analysis. Results revealed a significant rise in negative sentiment and a decline in positive sentiment following the public disclosure of the scandal on February 25, 2025, reflecting a deep erosion of public trust in Pertamina. Qualitative thematic analysis further identified a shift from neutral or positive discussions focused on service quality and innovation to emotionally charged critiques emphasizing betrayal, distrust and institutional failure. These findings highlight the value of integrating deep learning classification with qualitative insights to monitor real-time public opinion and institutional reputation. The study underscores the critical need for transparency and effective communication strategies during reputational crises to rebuild public confidence. Limitations include the focus on a single social media platform, suggesting future research should incorporate cross-platform and multilingual analyses. Practically, this research offers actionable insights for corporate crisis management and contributes to understanding social media’s role in shaping public trust and accountability in the digital age
Estimation of Blood Glucose Levels Using a Non- Invasive Infrared-Based Optical Sensor: A Pilot Study
Full text linkBlood glucose measurement is critical to diabetes management and prevents chronic complications such as neuropathy, nephropathy, and retinopathy. However, current methods are invasive, uncomfortable, and costly. Although several non-invasive approaches have been explored, no commercially available device offers a simple, affordable, and user-friendly solution for non-invasive blood glucose estimation, particularly one suitable for self-measurement outside clinical settings. This underscores the need for practical and inclusive alternatives. This study aimed to develop and evaluate a blood glucose estimation device using infrared LEDs to measure light transmittance through the fingertip. The research was conducted in two stages: initial testing using glucose solutions with varying concentrations (0.02, 0.06, 0.10, and 0.20 g/ml) and added red dye (0.05, 0.10, and 0.15 ml) to validate the sensor's response, followed by direct validation on human fingers against commercial blood glucose test strip readings. The results showed a strong positive correlation between sensor output and glucose levels, with a Pearson correlation coefficient of r = 0.995. Using a regression-based calibration model, the system achieved a mean absolute error (MAE) of 1.63 mg/dL, and a root mean square error (RMSE) of 1.72 mg/dL. Cross-validation, such as Bland-Altman analysis and Clarke Error Grid, was conducted to verify model robustness. These preliminary results suggest that the developed system holds strong potential as a simple, affordable, and non-invasive tool for blood glucose self-monitoring, especially in resource-limited settings. However, further validation on larger, more diverse populations is necessary
Energy Efficiency Consumption of Building Planning Based on Indonesia Green Building Rules: A Case Study
Full text linkThis study evaluates the energy performance of the Tower B building in Riau Province, Indonesia, based on green building assessment criteria, with a focus on energy use efficiency. The initial design showed a total energy consumption of 2,255,923 kWh/year, corresponding to an Energy Use Intensity (EUI) of 95.46 kWh/m²/year, and earned a moderate energy score of 30 out of 46. Two operational strategies were proposed to address inefficiencies: (1) reducing the number of elevator units from six to four, and (2) implementing separate grouping of luminaires based on daylight exposure. Simulation results indicated that these strategies could reduce energy consumption in the elevator system by 33% and in the lighting system by 13%, without compromising user service standards. The combined application of both measures reduced total building energy use to 1,949,861 kWh/year (EUI 82.51 kWh/m²/year), yielding a 29% overall reduction and increasing the energy efficiency score to 38. While the results are promising, further research is needed to validate implementation through sensitivity analysis, behavioral modeling, and integration into regulatory frameworks. This study highlights the potential of adaptive operational strategies in optimizing building energy performance in tropical urban contexts
Bridge-Crane Hoist Mass Angular Oscillation
Full text linkThe oscillation of hoisting masses in bridge cranes significantly impacts system stability, resulting in adverse effects such as excessive dynamic loads, inaccurate positioning during loading and unloading, and overall instability. Therefore, the angular oscillation of the hoisting mass must be controlled in the design of bridge cranes. This paper presents a mathematical model for the analysis of the angular oscillation of a hoisting load, both during oscillation and trolley movement. The MATLAB/Simulink software was used to simulate the dynamic behavior of the model to gain insights into the oscillation angle and its limitations. The results substantiate the reliability of the proposed model for predicting the angular oscillation of hoisting loads in bridge cranes