This paper examines condition-based maintenance (CBM) of complex technical systems using aircraft engines as an example. CBM requires continuous vibration monitoring to predict the optimal moment for scheduled maintenance. At any given moment, an engine may exist in multiple possible states, only one of which is realized. This set of possible states represents the physical potential of the system. To determine the actual state, this study proposes using Harrington's desirability function, which establishes a relationship between linguistic assessments of an engine's technical condition and normalized numerical intervals.