Data-Driven Urban Safety: A CNN-Based Predictive Model for Manhole Hazard Detection

The safety of urban environments is a paramount concern, with manholes serving as critical components of infrastructure by regulating sewages but also susceptible to pose hazard when improperly maintained or obscured. The passing years have shown the level of threat an open manhole poses. In this research paper, we try to introduce a novel approach to address these concerns by leveraging Convolutional Neural Networks (CNNs) for real-time manhole status prediction. This research offers a practical and effective tool for authorities to promptly address hazardous conditions and ultimately minimizing the risks to the public.