Health Monitoring of Mining Conveyor Belt using RFID Sensors: A ML Based Approach

Cracks in coal mining conveyor belts are the budding causes for structural failure and consequent loss of revenue. In this thesis, a novel Machine Learning (ML) based Radio Frequency Identification (RFID) sensing mechanism is proposed and experimentally validated for crack detection of both static and dynamic mine conveyor belt. A wide variety of crack detection scenarios are produced by creating submillimeter cracks of many sizes and orientations and tested with several ML algorithms. Such a pioneering implementation of ML using both chipped and chipless RFID is a novel contribution to the concept of remote monitoring of mining industry.