Abstract
Objectives. The identification of deviations and incidents in the activities of enterprises is an integral part of the entire process of compliance with regulatory industry standards. In particular, these standards include good manufacturing practices (GMP) and the system of corrective and preventive actions (CAPA). The aim of the research is to automate the processes of identifying deviations and determining the probable causes of their occurrence in complex stochastic systems. Methods. Questions concerning the modeling of compliance processes with industry standards are studied on the example of the organisational management of pharmaceutical production. A method is proposed for the automated realisation of deviation from identification processes based on the processing of unstructured messages, the automatic generation of characteristic control parameters and a comparison of registered values with normative ones. The problem of detecting the primary cause of deviations is considered on the basis of the algorithm for processing the cause-effect relationships and probabilistic values of the relationship between the different defined groups of deviations. Results. Based on the proposed method of machine detection and identification of deviations, the information management system of CAPA procedures has been developed and successfully implemented at several enterprises within the pharmaceutical industry. The feature of the proposed method is the principle of system self-development based on the processing of “historical” data, which allows the probabilistic values of the relationship between the different groups of deviations to be dynamically calculated. This in turn allowed decisions on the implementation of changes according to the expert reports generated by the information system to be quickly and accurately taken by quality specialists. The generation of the electronic production dossier has significantly reduced the time of production protocol preparation and eliminated human factor errors. Conclusion. The full-cycle automation of CAPA procedure management allowed the enterprises to solve the primary task of continuous compliance with industry standards due to timely detection of deviations or deviation trends and promptly carrying out corrective and preventive actions to eliminate inconsistencies.