Potholes tend to pose a notable risk to road safety and infrastructure. Potholes cause nearly more than 2,000 accidents per year. This survey paper provides an understanding of the state-of-the-art techniques and advancements in pothole detection using the YOLO (You Only Look Once) deep learning framework. YOLO has emerged as a positive solution for automating pothole detection. It then digs into the YOLO framework, explaining its architecture and advantages for object detection tasks. YOLOv8 features deep learning architecture incorporating 186 layers suitable for Real-time object detection. It also features custom filtering and labeling techniques to enhance accuracy and reduce false positives. The survey also talks about the evaluation metrics commonly used to evaluate the performance of YOLO-based pothole detection systems, such as precision, recall, and F1-score. The model achieved 98% accuracy in detecting potholes (0.98 map value for 0.5 IoU).