Machine Learning Based Crowd Sourcing Approach to Identify Road Surface Quality with Mobile Phone Sensors

Abstract

Transportation is crucial for development of the civilization in the world. Transport supports a major contribution on economic growth and globalization. In this present world, People seek for comfort in transportation in vehicles. Considering the number of vehicles using man made roads and poor maintenance on damaged roads, road surface quality can be an issue. Poor road face quality can cause vehicle suspension damage, tire punches and steering miss alignments. Due to those vehicle damages, people tend to use quality road surfaces. This project’s main focus is to provide road surface quality monitoring system based on machine learning and using crowdsourcing approach. The system provides a road surface quality monitoring system for the user to take necessary precautions before travel on a poorly developed road. To achieve the objective, data was gathered from users using crowdsourcing method. With those gathered raw sensor data, those raw data will be smoothed by using filters. After that process the data will be applied to an IRI calculation. After that those calculated data will be trained under a machine learning algorithm. With classifying those data and analyzes them, the quality of the road surfaces will be delivered.