METHODOLOGY

This methodology outlines a comprehensive approach for predicting bridge

deflection and stress responses, and assessing potential damage for the Ayala Bridge
in Manila, specifically tailored for implementation using Microsoft Excel. This approach
leverages quantitative research principles, systematic data collection, and regression
analysis capabilities within Excel to ensure the long-term safety and performance of this
critical infrastructure. The use of Excel makes this methodology accessible for data
management and analysis without specialized statistical software. Given its historical
significance and crucial role in Manila's transportation network, as well as past issues
with corrosion and seismic vulnerability, a robust monitoring and prediction is essential
to ensure its structural integrity against traffic loads, environmental factors, and potential
damage.

Research Design
The study used quantitative research design in compute the statistical
relationship between input variables load conditions, materials deflection and stress of
Ayala Bridge The research design will use a quantitative, predictive approach to
establish statistical relationships between various factors and the bridge's structural
responses. This framework is designed for systematic numerical data collection and
analysis, which can be managed and processed efficiently within Microsoft Excel.

Independent Variables (Predictors): These will include various loading conditions

(e.g., traffic volume, estimated vehicle weights, types of vehicles), environmental factors
(e.g., daily average temperature, wind speed), and structural characteristics (e.g., age
of bridge components, observed corrosion levels). Data for these variables will be
organized into distinct columns in an Excel spreadsheet.

Dependent Variables (Outcomes): The primary dependent variables are bridge

deflection and stress. These measurements will be recorded and stored in separate
columns in the Excel dataset, corresponding to the independent variables.

Theoretical Framework: The framework assumes that deflection and stress are
functions of applied loads and environmental conditions. Excel's analytical tools will be
used to empirically test these relationships.

Data Collection and Processing

The data collection and processing strategy is essential for accurate predictive modeling.
This section details how data will be acquired and prepared for analysis using Microsoft Excel.
 The Data will be collected from various sources and consolidated into a single
Excel workbook with different sheets or tables for organization. The Structural
Parameters of Bridge Geometry and Material Properties: Reference existing
engineering documents, blueprints, and reports to extract key structural dimensions and
material specifications. This information can be static and entered into a dedicated
Excel sheet. The maintenance Records is reviewing the historical maintenance logs and
inspection reports to note any repairs, modifications, or observed structural issues.
Summarize relevant findings (e.g., severity of corrosion, date of last major repair) into
numerical or categorical variables in Excel.

The Environmental Factors such as Weather Data to obtain daily or weekly

average temperature and wind speed data from local weather stations. This time-series
data will be input into Excel, aligned with the dates of structural response
measurements.

The Loading Conditions of Traffic Monitoring of the direct traffic sensor data is
unavailable, conduct manual traffic counts during peak and off-peak hours, classifying
vehicles by type (e.g., light vehicles, heavy trucks) and estimating their average
weights. Alternatively, use historical traffic data if available. This data will be aggregated
(e.g., daily total load, average vehicle weight) and entered into Excel.

The Structural Response Data (Deflection and Stress) by Manual

Readings/Simplified Sensors if advanced SHM systems are not feasible, use simpler
measurement tools (e.g., dial gauges, optical leveling) for periodic deflection readings at
critical points. For stress, if direct strain gauge data is not available, estimations based
on deflection and known material properties might be considered, though this would
introduce more uncertainty. These readings will be meticulously recorded in Excel,
noting the date, time, and specific location of measurement.

Regression Analysis
Regression analysis will be performed using Excel's built-in "Data Analysis
ToolPak" to predict bridge deflection and stress, identify significant variables, and
assess potential damage.

Model Selection and Application

Various regression models will be considered, with the selection depending on the
characteristics of the data and the complexity of the relationships observed.

Multiple Linear Regression (MLR):This will be the baseline model to establish linear
relationships between deflection/stress and multiple independent variables (e.g., load,
temperature). It helps identify the direct impact of each variable.
Regularized Regression (Ridge, Lasso, Elastic Net):These techniques are particularly
useful for datasets with many predictors or multicollinearity. They help prevent
overfitting and perform variable selection by penalizing large regression coefficients.

Variable Identification

Identifying the most significant variables is crucial for understanding the bridge's
structural behavior and optimizing monitoring efforts.

The Statistical Significance Tests for linear models, p-values associated with regression
coefficients will indicate the statistical significance of each independent variable's
contribution to predicting deflection and stress.

Feature Importance Techniques: For more complex models (e.g., tree-based ensemble
methods), feature importance scores (e.g., permutation importance, Gini importance)
will quantify the relative influence of each variable.

Damage Response Analysis

The regression models will not only predict deflection and stress but also aid in
identifying potential damage.

Thresholding Predicted Values: Predicted deflection and stress values will be compared
against established serviceability and ultimate limit state thresholds defined by
engineering codes and bridge design specifications. Exceedance of these thresholds
will indicate potential damage or an increased risk of damage.

Anomaly Detection: Deviations between predicted and actual responses (residuals) that
fall outside statistically defined confidence intervals or exceed predetermined thresholds
can signal anomalous structural behavior, potentially indicating damage. Regression-
based methods can be used to investigate energy dissipation for health monitoring,
where abnormal changes in mechanical properties indicate damage.

Correlation with Damage Indicators: Regression models can be developed to directly

predict damage indicators (e.g., crack propagation rate, stiffness degradation) if
historical data on such damage is available alongside load and response data. This can
inform real-time damage assessment and model-based transfer learning for
sustainability.

Model Validation and Interpretation

The developed regression models will undergo rigorous validation to ensure their
reliability and predictive accuracy.
The Model Validation is use for Standard statistical metrics will be used for quantitative
evaluation. The R-squared (R²):To assess the proportion of variance in the dependent
variable explained by the independent variables. The Mean Absolute Error (MAE) is to
quantify the average magnitude of errors in a set of predictions, without considering
their direction. The Root Mean Squared Error (RMSE) to measure the average
magnitude of the errors, with higher penalty for larger errors. The Mean Absolute
Percentage Error (MAPE) to express prediction accuracy as a percentage. The Cross-
Validation Techniques such as K-fold cross-validation will be employed to assess the
model's generalization performance and robustness on unseen data, minimizing bias
from a single train-test split. The Residual Analysis plotting residuals against predicted
values or independent variables will help check for homoscedasticity, linearity, and
independence of errors, ensuring the assumptions of regression are met. The
Comparison with Physical Models/FEA Where possible, regression model predictions
will be compared with results from established finite element analysis (FEA) models or
other physical simulations to cross-validate their accuracy. Integration with CNN, time-
frequency analysis, and fine element analysis are part of advanced SHM strategies.

Interpretation of Results

The Coefficient Analysis for linear models, the magnitude and sign of the regression
coefficients will be interpreted to understand the direction and strength of the
relationship between each independent variable and bridge deflection/stress. The
Feature Importance is to identified significant variables from feature importance
techniques will highlight the most influential factors affecting the Ayala Bridge's
structural response, guiding future monitoring and maintenance efforts. The Predictive
Insights to validated models will provide quantitative predictions of deflection and stress
under various future loading and environmental scenarios, allowing for proactive risk
management. Damage Assessment to Interpretation will include identifying conditions
and thresholds that indicate a high probability of structural damage or accelerated
degradation, triggering alerts for inspection or intervention. This assists in effective data
preprocessing and ML applications in SHM for different bridge types.

Conclusion

This Excel-based methodology provides a practical and accessible framework for

conducting regression analysis on the Ayala Bridge. By systematically organizing and
analyzing data within Excel, engineers can gain valuable insights into the bridge's
structural behavior, predict responses under varying conditions, identify critical
influencing factors, and flag potential damage. While Excel's capabilities are more
limited than specialized statistical software, its widespread availability and ease of use
make it an effective tool for an initial or ongoing monitoring and analysis program,
ensuring the safety and longevity of the Ayala Bridge.
[Link]
3811/9/10/178)[[Link]]([Link]
Intelligent_Transportation_System_Traffic_Load_Monitoring_and_Remote_Sensing_Str
uctural_Health_Monitoring_of_Critical_Bridges_in_Metro_Manila_using_Multi_Modal_A
nalysis)

[Link]
Intelligent_Transportation_System_Traffic_Load_Monitoring_and_Remote_Sensing_Str
uctural_Health_Monitoring_of_Critical_Bridges_in_Metro_Manila_using_Multi_Modal_A
nalysis
[Link]

[Link]

[Link]
[Link]/html/2509.18106v1
[Link]
[Link]