DECENTRALIZED DATA ORGANIZATION AND MANAGEMENT SYSTEM FOR SMART CITY INFORMATION TECHNOLOGY PLATFORMS

Abstract

Context. Smart City functions as a multidomain social and technical system in which engineering and infrastructure domains, as well as social and service domains, are connected through shared data, operational dependencies, and distributed project processes carried out by distributed domain teams. Under crisis conditions, degradation of a critical infrastructure domain may propagate to adjacent domains and trigger cascading effects. At the same time, contemporary approaches to urban data management still insufficiently integrate decentralized domain responsibility, formal modeling of interdomain dependencies, and practical mechanisms for governed data interaction within distributed teams.Objective. The aim of this study is to develop and validate a decentralized data organization system for Smart City that integrates Data Mesh principles, fuzzy cognitive modeling of interdomain influences, and a metadata-oriented governance layer to support coordinated work with data products in distributed domain teams.Method. The proposed approach combines a domain-oriented data organization system, fuzzy cognitive maps for representing cause-and-effect interdomain dependencies, and a governance layer implemented using DataHub and Google BigQuery. Practical validation was carried out using the Electric Power Industry domain as a critical infrastructure domain with a high potential for cascading impact. Four open datasets were integrated into the domain layer, describing attacks on energy infrastructure, electricity supply restrictions, the statuses of electricity market participants, and indicators of the day-ahead market. On this basis, domain data products and an integral indicator of cascading-effect risk were implemented.Results. The study proposes a decentralized data organization model for Smart City and demonstrates its practical implementation for the electricity domain. A domain portfolio of critical data products was established; owners and semantic descriptions were defined; data lineage relationships and data contracts were formalized; and an analytical representation was implemented to calculate the integral indicator cascade_risk_score. The obtained results show that the proposed architecture supports transparent interaction with domain data products, improves the reproducibility of interdomain integrations, and enables the identification of periods of elevated cascading risk. The practical implementation also confirmed that the model is not limited to a single analytical repository and can be extended to heterogeneous data sources used by different Smart City domains.Conclusions. The results demonstrate that the integration of Data Mesh principles, fuzzy cognitive modeling, and metadataoriented governance provides a viable foundation for data organization in Smart City under conditions of distributed team interaction. The proposed model supports decentralized responsibility for data products, improves the coordination of distributed project processes, and strengthens the capacity for the proactive detection of cascading risks in critical urban infrastructure. Further research should focus on interdomain empirical validation, scaling to Smart Region scenarios, and the integration of large language modelbased agents through the Model Context Protocol for the automated analysis of data contracts, metadata, and operational anomalies