ROADRANGER: MOBILE APPLICATION FOR TRACK AND ROAD DAMAGE REPORTING USING MACHINE LEARNING

Abstract

This project develops a machine learning-based road damage reporting application that allows users to report damage in real-time with photos and descriptions. Reports are forwarded to authorities via a web-based administrative dashboard, enabling validation, categorization and prioritization of corrective actions. The automatic notification feature keeps users informed about report developments. Testing is carried out to ensure a friendly user interface, cross-device compatibility, realtime response and an effective notification system. Evaluation includes compatibility with TensorFlow Lite, camera quality, and internet connectivity. The results show the app works well across a wide range of devices and network conditions, providing a consistent and responsive user experience. This project is expected to increase the efficiency and responsiveness of road maintenance services and strengthen community involvement in reporting road damage. This application can be a model for other infrastructure reporting systems.