Forecasting Road Crash Trend Utilizing Recurrent Neural Network: A Time Series Analysis Approach in the City of Manila

Urban areas face significant challenges in achieving sustainable development, with road safety being a critical aspect of urban sustainability. In the City of Manila, there are a lot of factors that contribute to road crashes and road accidents, posing risks to all road users including drivers, commuters, and pedestrians. This research proposes a solution leveraging the Recurrent Neural Network (RNN) for predicting road crashes depending on the time and location where the road crash occurred. By analyzing historical road crash data alongside various parameters, such as the date and location of the crash, the RNN model aims to forecast potential accident hotspots. Through this predictive approach, urban planners and policymakers can proactively address road safety concerns by prioritizing developing and maintaining traffic control infrastructure and policies that will lessen road crashes. The findings of this study offer valuable insights into enhancing sustainable cities and communities by mitigating road accidents through data-driven interventions and modernization efforts in traffic management systems. The findings of the study can also contribute to fostering innovation through integrating advanced technologies such as recurrent neural networks in infrastructure planning and management industries.