Software Architecture Validation of a Radiation and Chemical Monitoring System for Liquid Radioactive Media

Abstract

This paper presents the implementation and validation of a software prototype for an automated radiation and chemical monitoring system for liquid radioactive media based on an event-driven data model. The proposed architecture incorporates an asynchronous messaging layer based on Apache Kafka, local buffering of measurement events, separate storage of primary and processed data in RAW and PROC repositories, and mechanisms for data integrity verification and processing traceability. The developed prototype preserves primary measurement data in an immutable form, supports repeated processing of historical data, and ensures the independence of the measurement subsystem from analytical components. To validate the proposed architecture, a testing methodology including unit, functional, and integration tests was developed. During validation, 108,000 measurement events were generated and processed. Test scenarios covered transport layer failures, data integrity verification, protection of the RAW repository, traceability of processing results, and reproducibility of data processing. The results confirmed compliance with the requirements for data preservation, integrity verification, immutability of primary data, traceability of processing results, reproducibility of processing, and isolation of the measurement subsystem. The proposed approach demonstrates its technical feasibility as a basis for developing automated radiation and chemical monitoring systems for nuclear facilities.