Skip to content
Log in

Similar papers

Open access Jul 2026

Statistical parameter estimation and probabilistic forecasting of cargo turnover at seaports grouped by sea basin using predictive models

The article presents the results of an analysis of cargo turnover in five sea basins of the Russian Federation for 2021–2024. The analysis of port and sea basin operations, carried out on the basis of available statistical data, makes it possible to identify trends in their future development, which depend on both coastal shipping and foreign economic relations. Forecasting cargo turnover using existing methods enables sea basins and individual ports to optimise cargo-handling operations, plan organisational and technical measures for the operation and maintenance of lifting and handling equipment, coordinate cargo flow directions, and determine the amount of cargo to be stored in warehouses. Managing these indicators makes it possible to plan the utilisation of warehouse areas at seaports, prepare vessel consignments in a timely manner, and avoid vessel downtime during cargo operations. The analysis and forecasting of various types of economic activity are carried out using expert assessment and various mathematical methods, which are probabilistic in nature. This study applies the Holt-Winters method, which provided the most accurate forecast based on the results of the analysis. Following the cargo turnover forecast for 2025 obtained using the Holt-Winters method, the predicted values were compared with the actual data. To verify the degree of agreement between the forecast and the actual values, an additional calculation was performed using a built-in Microsoft Excel function. The application of these forecasting methods makes it possible to analyse cargo turnover in sea basins in order to mitigate potential conflicts during cargo transshipment, support prompt decision-making, and ensure the rational use of port infrastructure within each sea basin. The comparison of the methods considered shows that the cargo turnover values obtained using these methods are close to the actual statistical data.

I. Zub, S. A. Fomichev, V. A. Fomichev · 0 citations
Review Aug 2026

An integrated machine-learning approach for classifying severe incident risk in maritime construction using weather and operational data

This study develops a data-driven severity-classification approach for maritime construction, relating equipment activity from incident narratives to localized hydro-meteorological conditions to complement static safety systems. The research analyzed 1,139 maritime construction incidents from OSHA databases (2015–2025). A rule-based Natural Language Processing module classified unstructured narratives into 15 distinct equipment categories. Additionally, an interface to a historical weather API reconstructed micro-climate conditions at the incident locations. Sixteen machine-learning algorithms were comprehensively compared for injury severity classification using chronological hold-outs, stratified cross-validation and feature-ablation. A Logistic Regression classifier provided a strong balance of discrimination and interpretability, achieving a cross-validated AUC of 89.2% and a hold-out AUC of 95.6% for post-incident classification based on narratives, while a pre-incident configuration using only prior operational factors achieved an AUC of 68.7%, with an F1-score of 95.4% and a Brier score of 0.062. The predictive signal originates primarily in the incident narrative; weather and employer history contributed modestly. A sensitivity analysis confirmed performance was robust to employer-history exclusion, and a localized wind-speed association near 30 km/h was cautiously identified. The approach offers a conceptual triage aid for site superintendents. If integrated into Construction Safety Management Systems, this logic could prioritize hazard reviews and inform daily planning, pending operational validation. The study integrates unstructured narratives with quantitative geospatial weather metrics in maritime construction. It reports a highly transparent classifier as an alternative to opaque models, offering an applied integration of established methods for safety-critical risk analysis.

Amr A. Mohy · 0 citations
Review Open access Aug 2026

Practitioner-Informed AI Decision Support for Maritime Accident-Type Risk in Korean Waters

In maritime accident prevention, it is important to identify not only high-risk sea areas but also which accident-types are most likely to occur there. This study combines survey responses from 826 Korea Coast Guard practitioners with 3856 maritime accidents mapped onto an H3 grid over Korean territorial waters during 2021–2023, and proposes a practitioner-informed framework for predicting accident-type-specific risk. The survey showed limited use of quantitative, standardized accident risk criteria but high demand for AI-based prediction and area-level risk analysis. Practitioners’ perceived accident frequency differed substantially from the empirical accident distribution, whereas their prevention priorities aligned more closely with the actual pattern. Accordingly, this study treats the accident-type taxonomy not as a fixed prediction target but as a design variable of the label space for decision support. A two-stage framework first estimates accident occurrence at the H3 grid-time level and then classifies the accident-type conditional on occurrence. Comparing survey-aligned, data-aligned, union, sufficient-sample, and full administrative (7-class) framings under a common training protocol shows that accident-type organization creates trade-offs among field interpretability, coverage, class granularity, and predictive stability. The study thus reframes maritime accident prediction as an accident-type-specific decision-support problem-linking practitioner perception with empirical evidence.

Dayoung Kim, Won Choi, Seung Sim et al. · 0 citations
2026

On the Results of the Risk Assessment Study for Personnel of a Jacket-type Offshore Stationary Platform

The study is dedicated to professional risk assessment for personnel of Jacket-type offshore stationary platforms. The study provides an analysis and classification of typical accident scenarios specific to such facilities and includes a forecast assessment of the consequences for 140 employees. Six groups of scenarios have been considered: environmental impacts, fires and explosions, collisions with ships, helicopter crush, cargo drops, and destruction of substructure. Specific attention is paid to a disastrous scenario that involves substructure destruction, which may cause up to 140 fatalities. Individual and social risks have been calculated using two approaches: the quantitative assessment based on statistical data and the semi-quantitative index method (matrix approach). Calculation formulas and computation examples for the two most hazardous scenarios are provided: gas cloud explosion and destruction of the substructure. According to the quantitative calculation, the individual risk for personnel is 1.014·10–4 year, which is at the upper boundary of the ALARP (As Low As Reasonably Practicable) zone, whereas the scenario of substructure destruction makes the dominant contribution (9.5·10–5). Based on the comparative characteristics of the obtained results, the advantages and limitations of each method have been determined. The quantitative method ensures high accuracy and comparison with standards, but is labor-intensive, whereas the semi-quantitative method allows rapid risk ranking. The authors suggest a combined approach: the index method for hazard identification at the first stage and a detailed quantitative analysis of critical scenarios at the second stage, which enables efficient safety and platform structure integrity management.

I. Starokon · 0 citations
Open access Aug 2026

Traffic Accident Severity Prediction Based on Multi-Model Comparison

Traffic accident severity prediction is important for improving traffic safety management and supporting risk assessment and prevention. However, current studies have several limitations with respect to the systematic approach for comparing models and the comprehensiveness of evaluation metrics. This paper aims to systematically assess the overall performance of various machine learning models on a unified experimental framework to predict the severity of accidents in a binary classification task. This study randomly sampled 500,000 records from the US Accidents dataset and used them as the experimental sample. Following data cleaning, missing value handling, feature engineering, and categorical variable encoding, a comparison of four representative models was performed: Logistic Regression (LR), Random Forest (RF), XGBoost and Multi-Layer Perceptron (MLP). With the imbalanced nature of the data, this paper evaluates model performance using accuracy, AUC, ROC curves, and confusion matrices to assess the overall classification performance of different models and their ability to identify the severe accident class. The results show that the tree-based ensemble models Random Forest and XGBoost outperform Logistic Regression and MLP in terms of overall predictive performance, with XGBoost exhibiting better overall performance. Variables that contribute significantly to the model’s predictions include traffic control facilities, time factors, spatial location, and accident-affected distance. The results indicate that machine learning techniques could be useful in traffic accident severity prediction and could be a reference for traffic safety risk assessment. Meanwhile, class imbalance, feature interpretability and model generalizability are challenges that need further investigation and resolution.

Zi-Yu Cao · 0 citations
Open access Jul 2026

AI Enabled Risk Management in International Logistics Networks

International logistics networks are increasingly exposed to operational, financial, geopolitical, customs, cyber, and environmental risks. Effective risk management is therefore essential for reducing shipment delays, controlling logistics costs, and improving supply chain reliability. This study aims to examine how artificial intelligence-enabled classification models can support risk prediction in international logistics networks. For this research, a primary dataset was developed containing 3,000 shipment records with 30 variables, including shipment route, transport mode, cargo value, cargo weight, transit time, delay days, weather risk, political risk, customs risk, port congestion, cyber risk, insurance cost, total logistics cost, and actual risk level. The study applied supervised machine learning methods using a 70% training, 15% validation, and 15% testing split. Three AI-based classifiers—Random Forest, XGBoost, and LightGBM—were used to classify shipment risk into Critical, High, and Medium levels. Model performance was evaluated using accuracy, precision, recall, F1-score, and classification reports. The results show that Random Forest and XGBoost achieved the highest accuracy of 91%, while LightGBM achieved 90% accuracy. Random Forest and XGBoost performed strongly in identifying Critical risk shipments, both reaching an F1-score of 0.97 for the Critical class. LightGBM showed better recall for Medium risk shipments, indicating stronger detection of minority risk cases. Overall, the findings suggest that AI-enabled models can effectively support risk prediction and decision-making in international logistics networks. The study concludes that machine learning can improve proactive risk management, reduce uncertainty, and enhance logistics network resilience.

Purnima Tripura · 0 citations