A BigData/Machine Learning system for Monitoring Road Surface Condition

Preet Patel, Tiwaloluwa Ojo, Faraaz Mohsin, Janajan Jeyabalan, Waleed El Alawi, George Daoud, Mohamed El-Darieby

Abstract: Road surface condition and deterioration are crucial factors to monitor as it affects the safety and comfort of travellers. Currently, road surface inspection is conducted mostly with manual methods that require expensive machinery. Many municipalities lack the resources to conduct such inspections thoroughly and frequently. Collection and maintenance of updated data of road surface damage and conditions is crucial in making decisions regarding road maintenance. This paper leverages machine learning to create a cloud-based application that can detect different road conditions and damages. We describe the functional and non-functional requirements and discuss the detailed design of the software system developed and method of implementation for each component of the system. The paper also covers the integration and unit tests designed to test each component of the system and provides the results of these tests. The literature also provides the results for the acceptance tests and validates if the product developed meets the defined business requirements. Lastly, The paper discusses ethical and safety considerations.