The mining industry plays a significant role in societal growth globally. Mining operation is a complex process that often includes exploration, extraction, processing, and reclamation. The processing phase is intended to extract the mineral of interest, e.g. iron, from the raw material provided in the extraction phase. In iron mining, the processing phase often consists of crushing, grinding, separation, and filtration. The filtration plant is needed to remove the excess water from the fine-grained mixture of iron ore and water. Filtration is needed to maintain the moisture content and to produce a sandy iron ore mass. A filtration machine consists of multiple components that ensure effective operation to separate liquids from the mixture. These components are susceptible to failure, which, if undetected can affect the machine operation and may cause financial losses. Currently, a periodic visual inspection strategy is applied for health monitoring and assessment of the filtration equipment. The existing inspection strategy provides capabilities for diagnostics and failure detection. However, this inspection strategy has shortcomings, such as the insufficient capability to provide prognoses on the asset health, the estimation of the remaining useful life of the asset, the human factors, and the associated human errors. Therefore, the objective of this paper is to propose an approach for prognostic and health management of drum filters in mining production. This paper utilizes deep learning methods to identify failures in critical components. By leveraging image data and AI, this paper aims to enable early fault detection, ultimately improving the availability of the machine and lowers the maintenance costs.