Raman spectroscopy offers an attractive solution for monitoring key process parameters and predictive modelling in cell culture processes using transgenic Chinese hamster ovary (CHO) cells. Frequent in-line measurements offer the potential for advanced control strategies. However, an erroneous value created by analytical signal noise is a significant issue that can affect process controls negatively. One such challenge is to differentiate the signal reflecting process changes, ranging from random to gross error, in a timely manner so the process control system can respond to these changes and maintain adequate control.