ROADIO: A system for automated estimation and visualization of crowdsourced road surface quality using mobile phones

Abstract

Road infrastructure is crucial for economic growth as it facilitates the transportation of goods, services and people. However, poor road conditions can negatively impact the economy and hinder the timely delivery of goods and services. In Ghana, limited quality road infrastructure remains a challenge, and reliable information on road systems is needed to make informed decisions on infrastructure development. Mobile technology presents a promising opportunity to develop participatory sensing systems for collecting and processing road data. This study aims to further the development of a system for gathering, processing, and displaying data on road surface quality using crowdsourced sensory data from mobile devices in Ghana. The study seeks to create and efficient and cost-effective system that can significantly improve the state of road infrastructure in Ghana by leveraging the advancements in mobile technology and participatory sensing. We designed and implemented a three-tiered architecture software that enables large-scale detection of road surface quality through mobile phones. The system accommodates a machine learning algorithm for classifying road readings from accelerometers, and a service to address the problem of aggregating crowdsourced data which has not been effectively handled in previous research in the Ghanaian context. The classified road surface data is then visualized on a web interface that integrates a Google Maps API.