Electrical survey techniques for corrosion assessment in steel structures: case studies, equipment analysis, and practical applications

Introduction: the corrosion of steel structural components is one of the key pathologies at the level of civil engineering since it represents serious risks to the resistance and life of infrastructures. Objectives: this research aims to investigate the practical effectiveness of electrical surveys as a non-destructive technique for the diagnosis of corrosion of metal structures, especially in environments where the system becomes highly susceptible, such as bridges, tanks, pipes, and reinforced concrete elements. Methodology: the study was conducted through a systematic review of case studies and comparing experimental data with different survey configurations (Wenner, Schlumberger, dipole-dipole). Results: an analysis of electrical resistivity was conducted by environmental variables (humidity, salinity, chloride content), descriptive statistics and correlational analysis. The study found that 85% of the cases, with resistivity <100 Ω·m, presented conditions that were likely to be active corrosion. Similarly, we found an inverse correlation (r = -0.73) between resistivity and structural deterioration. The results indicated that the accuracy used was better in homogeneous soils, although it was also shown to fail in heterogeneous environments. Conclusions: it is suggested that electrical surveys can be an effective instrument to establish risk areas, however, if they are combined with techniques such as half-cell potential or impedance spectroscopy. This method allows for the development of planned maintenance plans, and this optimization of resources can improve and extend the life of the infrastructure. General area of study: Civil Engineering. Specific area of study: Metal structure. Type of article: bibliographic review.

Marlon Ariel Iñamagua Cuenca, Juan Sebastián Maldonado Noboa

6-26

Participation in complementary activities and academic engagement among students at Instituto Superior Tecnológico Vicente León

Introduction. Participation in complementary experiences can broaden students' academic and social integration, whereas academic engagement reflects behavioral, affective, and cognitive involvement in their education. Objective. To determine the relationship between participation in complementary activities and academic engagement among students at Instituto Superior Tecnológico Vicente León. Methodology. A quantitative, descriptive-correlational, non-experimental, cross-sectional study was conducted. For methodological and academic purposes, a 292-case database was analyzed using a 24-item Likert questionnaire. Internal consistency was estimated with Cronbach's alpha and associations with Spearman's rho. Results. Complementary participation reached a mean of 3.28 (SD = 0.89), with the medium level predominating (45.89%), while academic engagement reached 3.71 (SD = 0.72), with the high-level predominating (56.85%). The overall association was positive and significant (rho = 0.42; p < .001). Conclusion. Greater participation is associated with higher academic engagement; the association does not imply causality and should be confirmed with institutional field data. General Area of Study: Education. Specific area of study: Higher education and student participation. Type of study: Original article.

Henry Paúl Guanopatín Villalba, Raúl Oswaldo López Iglesias

27-42

Epidemiological response of black sigatoka disease (Pseudocercospora fijiensis) to different foliar calcium rates in banana (Musa AAA, Cavendish Subgroup) under commercial production conditions

Introduction. Black Sigatoka, caused by Pseudocercospora fijiensis, progressively reduces functional leaf area and is a major phytosanitary constraint in banana production. Objective. To evaluate the epidemiological response of the disease to different foliar calcium (Ca) rates used as a complement to commercial phytosanitary management. Methodology. A randomized complete block design with five treatments and four blocks was established in a commercial banana plantation in El Guabo, El Oro, Ecuador. All units received mancozeb at 1.4 kg/ha and differed in SOLEI CaO 20µ rates of 0, 1.0, 1.5, 2.0 and 2.5 L/ha. Disease dynamics were evaluated for 100 days through eight fortnightly assessments. Results. At L8, significant differences were detected for EE₃, EE₄, EE₅ and H+VLQ >5%. EE₅ showed the largest effect size (η² = 0.160). T3 (2.0 L/ha) recorded the lowest EE₅ (131.55), representing a 9.31% reduction relative to the control. Conclusion. The 2.0 L/ha rate showed the best performance within the evaluated range. Foliar Ca should be considered a complementary nutritional tool within integrated management and not a substitute for fungicidal control. General Area of Study: Agricultural Sciences. Specific area of study: Plant Pathology and Plant Epidemiology. Type of study: Original article.

Jefferson Estuardo Reyes Coronel, Luis Felipe Lata Tenesaca

43-58

Machine learning model for detecting cyberattacks in HTTP traffic in web environments

Introduction. The accelerated growth of web applications has increased HTTP traffic and has made this protocol one of the main targets for those seeking to compromise information security. Objective. To create a technological solution based on a machine learning model capable of detecting cyberattacks in HTTP traffic. Methodology. This was an applied study with a mixed approach. The HTTP DATASET CSIC 2010 was used; eight quantitative features were extracted, and Random Forest, SVM, and XGBoost were evaluated through accuracy, precision, recall, and F1-score, using grid search and five-fold stratified cross-validation. Results. XGBoost achieved the best overall performance, with 91.1% accuracy and an F1-score of 0.888, followed by Random Forest, with 90.7% accuracy and an F1-score of 0.884. SVM achieved a recall of 0.627 and failed to detect many attacks. Conclusion. Random Forest and XGBoost demonstrated the ability to identify anomalous patterns without relying on predefined signatures and constitute safer alternatives than SVM for a future production detection system. The findings support the viability of the proposed machine learning approach for strengthening cybersecurity in web environments. Future validation should use institutional HTTP traffic from the Technical University of Manabí and assess model performance under real operating conditions. General area of study: Engineering and technology. Specific area of study: Cybersecurity and machine learning. Type of study: Original article.

Jonathan Eduardo Mero Zambrano, Gabriel Eduardo Morejón López, Wendy Viviana Obregón Martínez

59-82