PATHLY– Smart Route Safety Recommender (Graduation Project)

Graduation project that involved building a mobile route planner that predicts the safest path based on real-time sensor data classification. Trained a DSC-BiLSTM model using both raw and FFT-transformed input to detect road types (potholes, speed bumps, etc.). Also, deployed a mobile-friendly backend interface that predicts the condition of each segment in the road based on its coordinates from the dataset.