The manufacturing sector is heavily influenced by artificial intelligence-based technologies with the extraordinary increases in computational power and data volumes. It has been reported that 35% of US manufacturers are currently collecting data from sensors for manufacturing processes enhancement. Nevertheless, many are still struggling to achieve the 'Industry 4.0', which aims to achieve nearly 50% reduction in maintenance cost and total machine downtime by proper health management. For increasing productivity and reducing operating costs, a central challenge lies in the detection of faults or wearing parts in machining operations. Here we propose a data-driven, end-to-end framework for monitoring of manufacturing systems. This framework, derived from deep learning techniques, evaluates fused sensory measurements to detect and even predict faults and wearing conditions. This work exploits the predictive power of deep learning to extract hidden degradation features from noisy data. We demonstrate the proposed framework on several representative experimental manufacturing datasets drawn from a wide variety of applications, ranging from mechanical to electrical systems. Results reveal that the framework performs well in all benchmark applications examined and can be applied in diverse contexts, indicating its potential for use as a critical corner stone in smart manufacturing.