Road Condition Monitoring Using Axle-Based Acceleration Method and K-Means Clustering Algorithm

Road deterioration remains a major setback if it is not considered significant especially in Pakistan which causes human casualties and massive financial losses. Hence, road condition monitoring is helpful to ensure comfort and safety to drivers on road. The aim of this project is to design an effective low cost Axle-Based Acceleration (ABA) system for road condition monitoring to restrain road accidents and vehicle damages on roads. The prototype consists of NodeMCU (with built-in WIFI) and ADXL335 accelerometer is deployed on the wheel axle of vehicle to follow ABA method. However, the android application mapped with Google Map is designed to collect the location coordinates of the vehicle. This data using Hashmap is then continuously sent to the Firebase database using WIFI module for analysis. ABA methodology easily detects short irregularities efficiently as compared to the sensors mounted on other positions which are unable to capture low frequency vibrations. Firstly, the ABA method is validated in this paper which shows that it is 12.849 % more efficient than the Non-ABA methodology. Thereafter, the collected data (Acceleration, Speed and GPS coordinates of vehicle) on Firebase is analyzed using K-Means clustering algorithm for prediction of faulty location coordinates. Those faulty coordinates are then entered into the database so that the road authorities can fix them before the road condition becomes worst.