Definition
Classification of variables affecting the process into constant, noise and experimental variables: C = constant (without change over time) / N = noise (with change over time) / X = experimental variable; X does not use in VRK.
In detail
Variable Classification (CNX) is a method for classifying variables affecting a process into Constants (C), Noise (N), and Experimental variables (X). Constants do not change over time; Noise changes over time but is uncontrollable; Experimental variables are controllable and adjustable. In Variation Reduction Kaizen (VRK) weeks, X is not used because the kaizen week focuses on identifying and reducing noise, not experimental optimization. It helps teams focus on key variables, distinguishing controllable from uncontrollable factors. Typically, a CNX table is formed, listing variable classifications and conversion actions.
Example in practice
In an injection molding workshop, product dimensions fluctuate significantly. The team uses CNX classification: mold temperature is a constant, ambient humidity is noise, and injection pressure is the experimental variable. However, during the improvement week, the team does not adjust injection pressure but instead controls humidity as a constant, such as installing a dehumidifier to reduce fluctuation. The final output is a CNX table clarifying the handling of each variable, stabilizing the production process.