In smart manufacturing, companies continuously gather large volumes of data from machine sensors. Extracting value from this data requires advanced analytical methods to integrate, analyze, and predict potential issues, which is where Prognostic and Health Management (PHM) comes into play. PHM is a comprehensive framework that uses data mining, big data analytics, and artificial intelligence (AI) to perform condition monitoring, fault diagnosis, failure prediction, and lifespan tracking of industrial machines. The goal of PHM is to improve machine performance, reduce downtime, and optimize maintenance strategies.In this research, unsupervised learning techniques are employed to process the vast amounts of raw, unlabeled data commonly found in industrial environments. These techniques transform raw data into machine health scores, allowing real-time monitoring of machine conditions. By visualizing health scores of each machine, the system helps on-site personnel rapidly identify anomalies and determine whether predictive maintenance is required. This shifts the focus from conventional preventive maintenance, which is often time-based, to a more efficient predictive maintenance approach, which is driven by actual machine performance data. Moreover, this research integrates optimization algorithms to enhance the configuration of air compressors, a critical component in many industrial systems. By optimizing the operation and configuration of these compressors, the study aims to maximize their efficiency, reduce energy consumption, and ensure consistent production line performance. This combined approach not only supports predictive maintenance but also leads to significant cost savings and improved overall system reliability, making it an essential tool in modern smart manufacturing environments.