MUHAFIZ: IoT-Based Track Recording Vehicle for the Damage Analysis of the Railway Track
Ali Akbar Shah, Naveed Anwar Bhatti, Kapal Dev, Bhawani Shankar Chowdhry
Abstract
Fault diagnosis plays a major role in railway condition monitoring, as early diagnosis of the emerging faults can save valuable time, reduce maintenance costs and, most significantly, help save people's lives. However, the conventional data-driven methods used to diagnose track faults, especially in underdeveloped countries, use push trolley/train-based track recording vehicles (TRV) that rely heavily on manual extraction of track data. It is a very demanding process and significantly affects the final results due to its reliance on human judgment in assessing track conditions and its suboptimal performance. In contrast, with the advent of IoT-based smart inertial measurement units, the data-driven fault diagnosis became a core component in the smart industrial automation safety system. We proposed, Muhafiz, a prototype that is an automated and portable TRV with a novel design based on axle-based acceleration methodology for rail track fault diagnosis. Our contribution concluded, based on site-specific experimentation, that Muhafiz is 87% more efficient than the traditional push trolley-based TRV mechanism.