Data Mining for Speed Bump Detection from Car Wheels and GPS Sensors using Random Forest

Abstract

Wireless Sensor Network is several sensor nodes that form a network capable of communicating with each other wirelessly. Data mining is useful for finding a pattern from a large amount of data, for driving condition monitoring can be obtained through Controller Area Network (CAN)-bus data. Vehicle sensors for this research provide signal data, including wheel speed and GPS data. Random Forest is used for this project because of it easy to use and understand. Random Forest Classification is one of the popular methods of data mining. Sensors, applications, and users who want access the vehicle information are integrated to communicate with each other using Mobius IoT platform. The accuracy of Random Forest Classifier obtained through this research is 80.9%.