A New Era in Railway Track Inspection: Drone based Image processing integrated with IoT

In contemporary times, with the increasing demand for rail transportation, the preservation of the integrity and overall condition of railways holds paramount importance. This is because any damage to these routes can lead to irreparable disasters. Methods such as human inspecting, due to the drawbacks it entails, is not optimal nowadays. On the other hand, more modern approaches, such as installing cameras on trains and carts, face limitations like blocking the path and being dependent on train movement, which constrains the freedom of inspector movement. In the proposed method of this paper, the assessment of railway health is conducted through the processing of video footage captured by a camera mounted on a quadcopter, also known as a drone. In this method, the drone employs LiDAR technology for environmental scanning, which is performed to maintain a consistent altitude and protect the drone from external hazards. Drone-based image processing technique, combined with an understanding of the operational environment, allows for a considerably more cost-effective, accurate, and, of course, quicker assessment of the railway tracks health. However, factors such as lack of inter-drone communication for enhanced coordination, absence of data archiving, unavailability of missions history, and periodic health status reporting are considered weaknesses that challenge the scalability of this method on a larger scale with various complexities. Some of the key points of this study is about accessing real-time data about each drone during the mission. Moreover, within the dedicated profile of each drone developed within the central control system, detailed information on all completed missions, identified damaged points on railways, and periodic health reports for each drone are provided. With this process, the management of railway line health and safety will be significantly more straightforward and adaptable through the use of this new method.