International Journal of Engineering Technology and Natural Sciences

Identifying Safety Risk Sources in Bridges Construction: A Literature Review

Chizanatul Chusnia, M. Rendy Wibawa, Pungky Dharma Saputra — Sun, 28 Dec 2025 20:06:48 +0700

The construction sector, particularly bridge construction, plays a vital role in the development of national infrastructure but also involves significant safety risks that may lead to various losses. To address these risks, effective risk management is required to identify potential hazards and prevent workplace accidents. This study aims to identify the sources of safety risks in bridge construction projects through a systematic literature review. A total of 100 relevant studies were collected and analyzed using the PRISMA approach, while VOS viewer was employed to map and visualize the relationships between risk factors. The findings indicate that six main sources of risk can occur in bridge construction projects, with human-related risks being the most frequent, followed by environmental and managerial factors. These results highlight the importance of systematically identifying safety risk sources to strengthen preventive measures and reduce the likelihood of accidents in construction projects. This study relies exclusively on secondary data derived from published literature, which may restrict its ability to comprehensively represent context-specific, dynamic, or emerging safety risks encountered in actual bridge construction practices. Future research should incorporate empirical investigations, such as field observations, surveys, or in-depth case studies, to validate the identified risk sources and to develop more robust, context-sensitive risk management frameworks for bridge construction projects.

Mechanistic Insights and Optimization of Phytol Recovery from Acacia Auriculiformis Leaves Using Zinc Chloride Catalysis

Abubakar Ali, Haruna Ibrahim — Mon, 01 Dec 2025 00:00:00 +0700

This study explores a sustainable and environmentally friendly approach for phytol production from Acacia auriculiformis leaves, an underutilized lignocellulosic biomass, in response to the growing global demand for renewable bio-based chemicals. The research aims to optimize phytol extraction through zinc chloride–catalyzed thermal hydrolysis under mild reaction conditions while maintaining high selectivity and yield. The method employs ZnCl₂ as a Lewis acid catalyst to facilitate chlorophyll cleavage, with systematic variation of reaction temperature (40–80 °C) and catalyst loading (0.5–1.5% w/w) to determine optimal processing conditions. The highest phytol yield, 646.26 mg/g (13.14%), was obtained at 50 °C with 0.5% ZnCl₂, exceeding yields reported for other plant sources and conventional extraction techniques. Product characterization using gas chromatography–mass spectrometry (GC-MS) confirmed phytol as the dominant compound, accompanied by minor hydrolysis by-products. Mechanistic analysis revealed that yield variations were influenced by the balance between phytol formation and thermal degradation pathways under different catalytic and temperature conditions. These findings demonstrate the strong potential of A. auriculiformis leaves as a renewable feedstock for phytol production and highlight the effectiveness of ZnCl₂-assisted hydrolysis in supporting circular bio-economy and green chemistry principles. However, further studies are recommended to evaluate process scalability, economic feasibility, and environmental impacts to support industrial-level application.

Human Intruder Detection System (IDS) for Restricted Security Area: A Systematic Literature Review

Yadhurani Dewi Amritha, I Made Yogaswara Dipta — Wed, 31 Dec 2025 19:44:31 +0700

Ensuring security in sensitive areas such as airports, military bases, and nuclear facilities is critical to prevent unauthorized access. Traditional reliance on security personnel is often inefficient and insufficient for continuous monitoring. Intruder Detection Systems (IDS), which utilize devices or sensors to detect unauthorized entry, have emerged as essential tools for safeguarding high-security environments. However, there is a lack of comprehensive understanding that systematically synthesizes existing research on human intruder detection. This study aims to conduct a systematic literature review (SLR) on human IDS to provide a structured overview of current methodologies, technologies, and challenges in the field. Using established SLR protocols, relevant studies were collected, analyzed, and categorized to identify prevailing trends and gaps. The results highlight various object detection techniques and their effectiveness in real-world security applications. Despite the advances, challenges such as limited environmental adaptability and real-time accuracy remain. The findings of this review offer valuable insights for professionals and future researchers, guiding the development of more robust and efficient human intruder detection solutions.

Developing Mie Lidi X Packaging Concept Using Kansei Engineering Method

Wed, 31 Dec 2025 19:54:14 +0700

Packaging design plays a critical role in influencing consumer purchasing decisions, as it functions not only as a protective medium but also as a key marketing instrument. Although having an established market, Mie Lidi X currently employs overly simple packaging that offers limited visual appeal and inadequate product protection. This study aims to develop an improved packaging design concept that aligns with consumer preferences and emotional responses. A Kansei Engineering approach was employed to capture consumers’ affective perceptions of packaging, supported by Principal Component Analysis (PCA) to identify dominant design dimensions. Data were collected through questionnaires distributed to 31 respondents, of whom 74.2% indicated that packaging design is very important, while the remainder considered it important. In addition, observational analysis identified 57 packaging samples and 37 Kansei words relevant to snack packaging. PCA results revealed four principal components with eigenvalues greater than one and a cumulative explained variance exceeding 80%. These components were subsequently interpreted with the assistance of expert panelists. The resulting design concepts were identified as PC1 “Simple–Vintage,” PC2 “Attractive–Fragile,” PC3 “Strong–Modern,” and PC4 “Sustainable–Irrelevant.” The findings provide a conceptual foundation for developing packaging designs that better reflect consumer preferences. However, this study is limited to the identification of design concepts and does not yet translate these concepts into specific visual or structural design elements. Future research is therefore recommended to operationalize the identified concepts into concrete packaging attributes and to evaluate their impact on consumer behavior and product performance.

Optimizing Online Shopping Experience: The Role of AI Chatbots and ‘Like’ Features in Enhancing User E-commerce Engagement and Purchasing Decisions

Muhammad Rifai, Wahyu Andhyka Kusuma — Wed, 31 Dec 2025 20:01:08 +0700

In the midst of fierce e-commerce competition, user experience (UX) has become a crucial factor for success. However, users often face challenges such as slow responses to product inquiries and a lack of personalization, which can hinder purchasing decisions. This research aims to analyze and prove how optimizing UX through the implementation of an AI Chatbot and a ‘like’ feature can significantly increase user engagement and encourage purchasing decisions. The main focus is to design, test, and validate the effectiveness of these two interactive features as a solution to the identified problems. Using a Design Thinking approach, this study began with qualitative data collection through interviews to validate user needs. Based on the initial findings, an initial prototype (version A) and an optimized prototype (version B) were developed. Subsequently, the A/B testing method was used to quantitatively compare the two prototype versions to measure the increase in user acceptance after the optimization process. The results of the initial data analysis show that users perceive the ‘like’ feature as highly influential (80%) and the chatbot as positively influential (60%) on their purchasing decisions. The results of the A/B testing then confirmed that the optimized prototype (version B) received significantly higher ratings, especially in the ‘Very Good’ category, compared to the initial version, which validates the success of the design optimization process. In conclusion, the implementation of an AI Chatbot and a ‘like’ feature, designed through an iterative process, proved effective in creating a more efficient, personal, and satisfying shopping experience.

Implementation of Artificial Neural Network with Particle Swarm Optimization Algorithm for Financial Distress Prediction of Private Banks in Indonesia

Muhammad Alfin, Dafa Rifqi Firdianto; Noviyanti Santoso — Wed, 31 Dec 2025 21:14:44 +0700

Banking stability, particularly the risk of financial distress in private commercial banks, remains a critical issue that requires accurate and reliable prediction models. This study aims to analyze the characteristics of financial distress in Indonesian private commercial banks and to evaluate the effectiveness of Artificial Neural Networks (ANN) and ANN optimized with Particle Swarm Optimization (ANN-PSO) in predicting financial distress. Using financial data from 59 private commercial banks over the 2020–2023 period, this research employs five financial ratios as input variables and applies ANN and ANN-PSO models, with parameter selection conducted through a trial-and-error and optimization process. The results show that financial distress peaked in 2022–2023 with 32 distressed banks, while descriptive statistics indicate differences between distress and non-distress banks, including average NPLs of 1.40% versus 1.04%, ROA of 0.36% versus 0.75%, and LDR of 93.89% versus 92.39%, respectively. In predictive performance, both ANN and ANN-PSO achieved identical test accuracy of 95.74%, sensitivity of 93.75%, specificity of 96.77%, and an F1 score of 93.75%, although ANN-PSO demonstrated better model stability with lower training accuracy (98.40%) compared to ANN (99.47%), indicating reduced overfitting. Despite these promising results, this study is limited to a relatively short observation period and a fixed set of financial ratios; therefore, future research is recommended to incorporate longer time horizons, additional macroeconomic variables, and alternative optimization techniques to further enhance prediction robustness and generalizability.

Artificial Neural Network Based Evaluation of Wind Energy Potential for Small-Scale Renewable Power Generation in Wufeng, Taiwan

Haidar Rahman, Ahzami Fadilah Akbar — Wed, 31 Dec 2025 00:00:00 +0700

This study investigates the wind energy potential in the Wufeng area of Taichung, Taiwan, with the aim of supporting the development of small-scale renewable wind power generators. Specifically, it seeks to evaluate wind patterns and meteorological parameters over a three-year period and to identify the most accurate predictive model for wind speed and energy output. A quantitative research methodology was employed, analyzing weather data using multiple regression algorithms, including Linear Regression, Lasso Regression, Ridge Regression, Support Vector Regression (SVR), Dynamic Thermal Rating (DTR), and Artificial Neural Network (ANN). The performance of these models was compared through data training and testing, with the ANN demonstrating the highest predictive accuracy. Using this model, the maximum expected wind speed was determined to be 5.56 m/s, corresponding to a potential energy output of 992.57 watts over a one-week period, indicating that the region is suitable for small-scale wind power development. However, the study is limited by its reliance on short-term data, which may not capture seasonal variations, economic feasibility, or operational constraints of wind power systems. Therefore, future research should incorporate long-term wind monitoring, feasibility assessments, and pilot projects to evaluate the practical performance and reliability of small-scale wind turbines in the Wufeng region.

A Lean-Based Simulation Approach to Setup Changeover Improvement on the Flexo 8 Machine Using Single-Minute Exchange of Dies (SMED) Method

Nanda Dwi Febrian, Amalia Amalia — Wed, 31 Dec 2025 21:52:14 +0700

Production efficiency is a crucial element in enhancing the competitiveness of manufacturing industries. One common challenge is the lengthy setup changeover time, which leads to downtime and reduced productivity. This study aims to propose an improvement to the setup changeover process on the Flexo 8 machine at PT APP Purinusa Eka Persada-Semarang by implementing the Single-Minute Exchange of Die (SMED) method combined with Arena simulation software. The research was conducted through direct observation, data collection of setup times during January–March 2024, and analysis using the SMED approach, which includes separation of internal and external activities, conversion of internal to external activities, and simplification of setup tasks. Subsequently, a simulation model was developed using Arena software to compare the conditions before and after SMED implementation. The simulation results indicate that the SMED method successfully reduced the average setup changeover time from 52.67 minutes to 35.13 minutes, representing a 33.3% reduction. These findings confirm that the SMED approach can simplify setup processes, reduce downtime, and improve resource efficiency. The study recommends integrating SMED with simulation as an effective strategy for optimizing production processes. However, this research is limited by the exclusion of external activities in the simulation and the reliance on the accuracy of observational data. Future studies may expand the analysis to examine the impact of SMED on cost and product quality.

Application of Machine Learning for Classifying and Identifying Security Threats Using a Supervised Learning Algorithm Approach

Wed, 31 Dec 2025 23:02:38 +0700

The exponential growth of malicious web content has created an urgent demand for intelligent systems capable of accurately classifying cyber threats based on URL patterns. This study investigates the effectiveness of two widely used supervised learning algorithms, Random Forest and Naïve Bayes, in probabilistic classification tasks involving multiclass URL data. A synthetic dataset simulating 547,775 URLs was constructed to reflect realistic threat distribution: benign (65.74%), phishing (14.46%), defacement (14.81%), and malware (4.99%). Each instance was characterized by basic structural features such as length, dot count, HTTPS presence, and keyword indicators. To ensure fairness, both models were evaluated using identical stratified train-test splits across varying sample sizes, including a focused experiment on 15,000 and 100,000 entries. Results consistently revealed that both models exhibited high recall and precision only for the benign class, while failing entirely to detect minority classes. For Random Forest, precision and recall values reached 1.00 for benign URLs, yet dropped to 0.00 for phishing, defacement, and malware across all test sets. Naïve Bayes showed similar performance degradation, highlighting the severe impact of class imbalance and limited feature expressiveness. These findings emphasize the inadequacy of conventional classifiers in highly skewed, security-sensitive environments without preprocessing interventions. The study concludes that while Random Forest and Naïve Bayes offer computational simplicity, their default behavior is biased toward majority classes, rendering them unsuitable for detecting cyber threats without employing resampling techniques (e.g., SMOTE), cost-sensitive learning, or feature augmentation strategies. Future work will explore adaptive hybrid models with contextual features and deep learning frameworks to improve multiclass detection in real-world cybersecurity applications.

Green Architecture for Sustainable Hotel Design in Sosromenduran

Raden Zoland Bintang Permanaa, Endy Marlina, Desrina Ratriningsih — Thu, 01 Jan 2026 02:29:48 +0700

Yogyakarta is a major tourism destination undergoing rapid expansion of hotel and tourism infrastructure, resulting in increased environmental pressures such as greenhouse gas emissions, rising temperatures, and urban pollution. These impacts of global warming underscore the need for environmentally responsible development in the hospitality sector, positioning Green Architecture as a strategic approach to achieving sustainable tourism development. This study aims to explore the implementation of Green Architecture principles in hotel design as a means of reducing environmental impacts while meeting the growing demand for accommodation in Yogyakarta. The research employs a qualitative, design-based approach by integrating green architectural strategies into the planning and design process. Key strategies include optimizing building orientation and massing in response to solar patterns, maximizing natural ventilation and daylighting, integrating photovoltaic systems as renewable energy sources, and applying rainwater harvesting to supplement water supply and reduce groundwater exploitation. Landscape design and green spaces are also emphasized to improve microclimate conditions and thermal comfort. The results demonstrate that the proposed Green Architecture approach can enhance energy efficiency, reduce reliance on mechanical ventilation and artificial lighting, and improve indoor and outdoor environmental comfort. However, this study is limited to a conceptual design framework and does not include post-occupancy or operational performance evaluation. Future research is recommended to assess long-term energy and water savings and user comfort to strengthen the practical application of Green Architecture in hotel development.