Abstract
Management of procedures for corrective and preventive actions is one of the main tasks for enterprises in such areas as biotechnology, food and pharmaceutical production. In practice, the process of detecting and identifying nonconformities is fraught with a number of difficulties due to the random nature of their occurrence and the length of the process of identifying the sources that generate these discrepancies. There has been presented a step-by-step process of detecting and identifying inconsistencies, based on the analysis and formal description of the corrective and preventive procedures. A method for classifying messages on inconsistencies coming from various sources and machine definition of key parameters for controlling process stability has been proposed. The principle of self-development of the computer system is implemented, which allows dynamically updating key parameters and calculating their reference values on the basis of empirical data and generated scoring context-sensitive tables. In particular, data granularity indicators, such as "frequency" and "number of registrations" are considered as such parameters and a practical example with a step-by-step description of the calculation of their values is presented. An approach to the automation of computing processes is proposed, by additionally taking into account aggregated variables, which are criteria for estimating key parameters. Practical use of this method allows real-time identification of inconsistencies and promptly conducting corrective and preventive actions to eliminate detected inconsistencies.