The paper introduces a novel intelligent modeling system of a railway cargo delivery which combines queuing theory and station-level technological activities to model the manner in which re-handling and waiting processes produce delivery delays. The proposed model is in contrast to the available literature, which focuses more on routing or time management; it clearly connects processing of the stations, queue behavior, and reliability of the delivery in one decision system. When applied to a real-life railway route, the optimization of technological sequences is demonstrated to decrease delivery time and congestion rates significantly, as well as decrease the possibility of punishment in case of late deliveries. These findings show that the study is original in terms of the presentation of a data-driven and operationally based approach on the enhancement of railway freight performance. This study introduces a shipment-type-specific intelligent delivery model that integrates queuing theory with real station technological processes. Unlike existing approaches focused mainly on routing or average travel time, the proposed framework explicitly accounts for wagon processing sequences, re-handling operations, and delay-risk assessment. Validation on the Khamza–Bukhara corridor demonstrates a reduction in intermediate re-handlings from four to two and total delivery time from 68 h to 54 h, confirming the operational and economic effectiveness of the model.
Barotov et al. (Thu,) studied this question.