인천대학교 인디고클랫폼 인천대학교 기술지원허브

This paper proposes a real-time Ram Tilt variation monitoring system to detect equipment abnormalities and stabilize production in multi-stage cold forging processes. As equipment wear progresses, misalignment often occurs, leading to product defects and tool damage. The proposed system utilizes dual distance sensors to capture micro-variations in Ram Tilt during reciprocating motion and calculates the Ram Tilt deviation (mm) through real-time processing. Data collected over one year from eight forging machines were analyzed using SPC X̄-R charts, and abnormal control patterns were classified into three types: chronic instability, intermittent spike, and mean shift. The system was integrated with an MES to automatically stop equipment when deviations exceed thresholds, minimizing downtime and tool damage. Furthermore, the statistical independence between cycle time and Ram tilt (Pearson correlation coefficient r≈0) was verified, proving its reliability as a condition indicator. The results demonstrate improved manufacturing reliability, including a 98% reduction in defect rates, and reduced operator dependence, suggesting potential applications in smart factory environments with low-volume, high-mix production.