Corrosion Prediction in Steel Bridge Structures Using Artificial Neural Networks in Archipelagic Environments

Corrosion in steel bridge structures remains a major concern that can endanger the reliability and safety as well as shorten the service life of infrastructure, especially in regions with extreme environmental conditions, such as archipelago. Corrosion also plays a main contribution to sustainability issues anymore with green construction practices, as it enhances the clear strain on resources or with energy, resulting in materials in the repairing and the maintenance structures, rendering there so generation event or in addition time struggle. This study aims to cast light on forecasting of Bridget corrosion based on incorporating sustainability in usage of ANNs. A dataset having ecological parameters was used to develop the ANN model. Other dataset entries comprised material properties and historical corrosion records. Data was obtained from the Riau Islands region to obtain a representation of dynamic environmental conditions. The proposed ANN model demonstrated strong predictive capability, achieving a coefficient of determination (R²) of 0.91 and a mean squared error (MSE) of 0.045. These results indicate the model’s robustness and reliability in forecasting corrosion progression in steel bridge structures exposed to harsh environmental conditions. Based on true ANN modeling results, the most considerable factor affecting the 10.6% corrosion rate was Chloride Ion Concentration. Also, there are 8.7% of Sulphur dioxide (SO₂), and 7.2% of Nitrogen oxide (NOₓ). Additionally, the study highlights the potential application of sustainable materials and maintenance practices derived from ANN predictions to mitigate environmental concerns and minimize long-term maintenance costs. ANN also supports a Life Cycle Assessment to assess the environmental impact of the different maintenance scenarios. This enables decisions to be made with a consideration of technical and environmental implications at once.