Manufacturing Process Optimization Using Open Data and Different Analyses Methods

Machining process optimization involves selecting appropriate control variable (CV) settings to achieve desired evaluation variable (EV) outcomes. With the emergence of Open Data (OD) in smart manufacturing, machining optimization can now incorporate diverse CV-EV-centric data beyond local ones. This study investigates whether CV-EV-centric OD provides sufficient information, whether its analysis can yield actionable insights, and how suitable optimization methods are for OD-driven analysis. To explore this, Analysis of Variance (ANOVA) and Signal-to-Noise Ratio (SNR) were applied as conventional methods, while Possibility Distribution (PD) was used as a non-conventional method. The results indicate that PD offers an integrated solution, combining the strengths of ANOVA and SNR, for extracting actionable insights from the OD. It (PD) also enhances interpretability for some CV-EV-relationships and enables the formation of linguistic rules—an aspect not directly achievable through ANOVA or SNR alone. The findings suggest that combining conventional and non-conventional methods improve the analysis of OD-driven machining data, contributing to more structured optimization frameworks in smart manufacturing.