IoT System for Sensing Condition of Roads Using IMU Sensors

摘要

Paper presents a prototypic design and study of an IoT-based system to sense the condition of roads using sensor nodes installed inside the vehicles. Poor condition of roads is one of the major problem in the transportation system. A system that can identify poor condition of roads and make this data available in usable format can help in planning a better maintenance strategy by government bodies. It can also be part of navigation services to provide better route recommendations to the users. In the presented work, the system identifies the damages/irregularities on the roads like potholes, humps, mud holes, rough surfaces, etc, and marks them as red zones. The good condition smooth roads good for transportation are classified as green zones. This information about red zones and green zones is transmitted to a centralized cloud and is viewed in a usable format on an interactive GUI. In the presented work, a prototype of such a system is developed using IMU sensors - accelerometer and gyroscope. Supervised and Unsupervised learning algorithms are tested and deployed to carry out the classification. Papers briefly discuss the hardware prototype, data preprocessing techniques, classification algorithms, and their accuracy. Results obtained during real-time testing of the system are presented.