Real-Time Detection of Speed Breaker and Potholes Using Machine Learning for On-Road Drivers and Visually Impaired Persons

Well-maintained roads are vital to a country's economy as they provide essential transit routes. Potholes and speed breakers, on the other hand, seriously endanger drivers and reduce safety. Inadequate night vision, inaccurate speed breaker signage, and careless driving are among the causes of these risks. On uneven roads, vehicles with limited ground clearance are hazardous. To address this, we provide a tool that uses a Recurrent Neural Network (RNN) integrated into a machine learning method. With the help of an ESP32 GPS Speedometer, the device can identify potholes and bumps and sends the position information including speed, latitude, and longitude to a cloud database using WiFi. Additionally, it uses a speaker to deliver speech alerts. Furthermore, MATLAB's Bayesian Regularization improves suspension health and lessens driver fatigue. Our solution integrates voice alerts, GPS navigational systems, and RNNs to improve road safety.