Fatma Kutlu Gündoğdu

Early identification of risky driving behaviors is crucial for enhancing safety systems and advanced driver assistance technologies. A significant portion of traffic accidents, particularly under congested traffic conditions, arises from high-risk driving patterns that are not detected in time. This study presents a data-driven approach to identify and classify risky vehicles using the US-101 highway segment from the publicly available Next Generation Simulation (NGSIM) dataset provided by the U.S. Federal Highway Administration. In the first stage, risk indicators were extracted to quantify vehicle interactions and assess potential collision risks. These indicators take into account changes in vehicle speed, following distance, and braking behavior. Vehicles were then clustered using the Fuzzy C-Means (FCM) algorithm on the basis of similarities in the identified risk indicators. Subsequently, each cluster was labeled with an appropriate level of driving risk. Using the risk indicator values and the assigned risk labels of the vehicles, four supervised classification algorithms including XGBoost, Random Forest, Extra Trees, and LightGBM were employed and their performance was evaluated. The results demonstrate that the used risk indicators contribute significantly to the accurate classification of driving risk levels. This framework shows potential for supporting risk prediction modules in traffic safety management applications.

This study proposes a quantitative, expert-driven framework to identify and prioritize risks arising from the use of artificial intelligence (AI) in the military command-and-control cycle (C2). Recent operations, including the large-scale AI-supported strike campaign against Iran in 2026, illustrate how artificial intelligence is no longer a future capability but an active determinant of operational tempo, targeting precision, and intelligence-surveillance-reconnaissance throughput in real-world C2 cycles. Building on related literature, governance frameworks, and expert input, we define seven main risk categories and twenty-seven sub-risk criteria spanning data, algorithms, outputs, operational context, resources/infrastructure, security threats, and ethical-legal compliance. To incorporate both uncertainty and confidence in expert input, we employ a Spherical Fuzzy (SF) Z-number multi-criteria decision-making design: the SF-Z Best–Worst Method weights the main categories, while the SF-Z Analytic Hierarchy Process prioritizes the sub-risks. Results show that the highest priorities are concentrated on the security, resources and infrastructure, and data quality domains, while output-related risks receive the lowest relative importance, suggesting stronger manageability through human oversight and verification mechanisms. A robustness comparison using classical AHP indicates broadly consistent patterns, while the SF-Z approach yields more realistic prioritization by explicitly modeling uncertainty and confidence. As AI-enabled C2 proliferates beyond leading-edge militaries, the associated risks, spanning technical, operational, legal, and ethical dimensions, become a shared governance challenge. The framework supports risk-informed governance and capability development for artificial intelligence-enabled command-and-control by highlighting where mitigation investments are likely to produce the greatest operational value.

Precisely assessing driving danger is essential for various applications, including the advancement of autonomous driving systems and traffic engineering decisions. This study presents a driving risk analysis framework based on the Next-Generation Simulation (NGSIM) dataset. First, vehicles were classified into four risk classes using the Fuzzy C-Means algorithm using five key risk indicators. Subsequently, comprehensive driving behavior features representing vehicle movements were extracted and evaluated for both risk class prediction and driving behavior feature selection. A new driving risk score was developed using Spearman’s rho coefficient weights, which reflect the relationship of each risk indicator to risk levels. This score was observed to exhibit an increasing trend consistent with the sequential structure of the Fuzzy C-Means (FCM) clustering based on risk labels, thus confirming that it accurately reflects the labeling process. Furthermore, the findings show that the 26 key driving behavior features selected can predict the driving risk score developed using the XGBoost algorithm with over 85% accuracy. Moreover, feature importance analysis reveals that the following distances and inter-vehicle distance variability are particularly effective in determining driving risk. The study discusses the limitations of driving risk assessment based solely on vehicle dynamics and highlights the importance of developing enriched datasets that include multidimensional data sources such as environmental conditions, infrastructure features, traffic density, and autonomous vehicles in future risk prediction studies. Ultimately, this framework contributes to the development of safer and more efficient transportation systems, supporting environmental sustainability by reducing accident-related congestion and promoting resource-efficient traffic management.

The Autonomous Vehicle (AV) industry is constantly growing, thus analysing its perspectives is essential. However, for this analysis a sophisticated approach is necessary which considers the ambiguity of decision-makers, and different objectives and criteria related to stakeholders. In this paper a new model is proposed based on Decomposed Fuzzy Sets and the Best-Worst Method to deal with possible non-reciprocity of pairwise comparisons and different preferences of stakeholders in the AV industry. The main advantage of the model is that it is capable of considering optimistic and pessimistic attitudes along with the different objectives and criteria of the involved groups. The results show that users require short travel time, while operators, manufacturers and legislators expect mainly the increase of revenues from the AV implementation. Among the most important criteria, our analysis indicates the need of regulatory and safety issues are the most essential obstacles of expanding the AV industry. The new model can also be applied for evaluating the perspectives of other emerging technologies and industrial sectors.