Abstract— This article illustrates a systematic approach for predicting tool wear in machining process through Cyber-Physical System (CPS) architecture using simple electronic components such as personal computers and low-cost sensors. The proposed Cyber-Physical structure consists of 5 steps; smart connection, data to information, feature extraction, awareness of issues and self-adjustment. We tried to install a big data analysis technology into CPS architecture to catch the usual/unusual state of the cutting tool from the spindle power consumption changes. The excessive repetitions of grooving would bring the trend changing of power consumption. To facilitate the statistical analysis, the correlation coefficient R was calculated from the single regression analysis between two different cycles of time-series power consumption. The correlation coefficient R also had a strong relation with the condition changes of tool wear and would become a powerful tool to catch the usual/unusual state of the cutting tool in the proposed CPS architecture. The health information obtained from the system can be used for higher level of management of cutting tool based on the condition monitoring free from the schedule-based maintenance.