A method of correcting measurements of a chemical sensor used in an industrial facility. The method involves correcting for errors known to occur in the steady state and the dynamic state for specifically recognized situations. This method allows for correcting errors that occur due to deadtime, false zero measurements, and non-linear disturbances. The method combines automated measurement techniques and human know how to progressively learn and refine the accuracy of the corrections.