<![CDATA[This research aims to develop an artificial intelligence-based road performance prediction system to support smart infrastructure maintenance. Current road maintenance systems are still traditional and reactive, leading to infrastructure degradation and high repair costs. This study uses AI methods combining Artificial Neural Networks (ANN) and Long Short-Term Memory (LSTM) to analyze road condition data, traffic volume, and weather conditions. ANN is effective in detecting nonlinear patterns from statistical data, while LSTM excels in processing time-series data of historical road conditions. The system is designed using UML modeling and implements a relational database for storing road, traffic, weather, and prediction data. Based on the analysis, the proposed system successfully provides a predictive maintenance solution that is proactive rather than reactive. The system's performance demonstrates that AI-based predictions can extend road service life, optimize maintenance budget allocation, and minimize public service disruptions. However, prediction accuracy is still influenced by factors such as data quality and model parameter selection.]]>
Copyrights © 2024
