Aerospace Instrument-Making Annotation << Back Graph-Transport Method for Identifying Risk Factors and Assessing the Reliability of Complex Technical Systems using Weakly Structured Operational Data Konnikov E.A., Rodionov D.G. This paper develops a graph-transport method for identifying risk factors and assessing the reliability of complex technical systems from weakly structured operational data. The proposed GTBRN method is designed for multimodal operational datasets combining SCADA measurements, event logs, short alarm messages, and availability intervals. The novelty of the approach lies in representing each operational window as a multimodal distribution that jointly integrates a numerical regime profile, an event-semantic representation of weakly structured text, and the graph structure of subsystem interactions. An entropy-regularized transport deviation is used to compare the current state with normal and pre-event templates, while the final risk and reliability estimates are produced by a stable probabilistic head. Experimental validation was carried out on the open Fuhrlander wind turbine dataset for 2012-2014 and included comparisons with Cox PH, DSGRN, logistic regression, and Extra Trees. The results show that, at the 24-hour forecast horizon, the full GTBRN version achieves the best ROC AUC and Average Precision among the evaluated models, confirming its applicability to risk-based maintenance and early detection of hazardous operational windows. Keywords: weakly structured operational data; reliability of technical systems; risk factor identification; optimal transport; graph-based models; SCADA; alarm message analysis; risk assessment; risk-oriented monitoring. DOI: 10.25791/aviakosmos.8.2026.1575 Pp. 15-25.