“…They employed a random forest classifier to identify road potholes, achieving an accuracy of 95.7%, a precision of 88.5%, and a recall of 75.0%. Additionally, the study conducted by Zhou et al 18 focused on classifying the quality of manholes based on time and frequency domain features extracted from accelerometer and gyroscope data. They used a support vector machine to categorize manholes into three classes: good, average, and poor, which correspond to different levels of subsidence.…”