Path planning for autonomous underwater vehicles requires reliable and computationally efficient methods, particularly in cluttered environments. This work presents a comparative evaluation of representative approaches, including metaheuristic optimization methods (continuous genetic algorithm, particle swarm optimization, gray wolf optimizer, and Jaya), a sampling-based method (probabilistic roadmap with genetic refinement), a reactive strategy (artificial potential fields), and a control-based approach (model predictive control with control barrier functions). The algorithms are assessed in a controlled two-dimensional simulated workspace with randomly generated obstacles and systematically increasing obstacle density. Each configuration is evaluated across multiple independent trials using metrics such as success rate, path length, and convergence behavior. The effect of environmental disturbances is examined by analyzing particle swarm optimization under Gauss–Markov current models. The results show that performance depends strongly on the ability to preserve feasibility as obstacle density increases. The probabilistic roadmap with genetic refinement demonstrated the highest robustness, maintaining feasibility across all scenarios, while particle swarm optimization provided a strong balance between path quality and reliability in low-to-moderate clutter. The introduction of current disturbances led to reduced efficiency and consistency. Statistical analysis confirmed significant differences among methods, highlighting that rank-based superiority does not necessarily reflect practical robustness in constrained environments.

Wind direction nowcasting is crucial in various sectors, particularly for ensuring aviation operations and safety. In this context, the TELMo (Time-series Embeddings from Language Models) model, a sophisticated deep learning architecture, has been introduced in this work for enhanced wind-direction nowcasting. Developed by using three years of data from multiple stations in the complex terrain of an international airport, TELMo incorporates the horizontal $u$ (east–west) and $v$ (north–south) wind components to significantly reduce forecasting errors. On a day with high wind direction variability, TELMo achieved mean absolute error values of 5.66 for 2-min, 10.59 for 10-min, and 14.79 for 20-min forecasts, processed within a swift 9-ms/step timeframe. Standard degree-based analysis, in comparison, yielded lower performance, emphasizing the effectiveness of the $u$ and $v$ components. In contrast, a Vanilla neural network, representing a shallow-learning approach, underperformed in all analyses, highlighting the superiority of deep learning methodologies in wind direction nowcasting. TELMo is an efficient model, capable of accurately forecasting wind direction for air traffic operations, with an error less than 20° in 97.49% of the predictions, aligning with recommended international thresholds. This model design enables its applicability across various geographical locations, making it a versatile tool in global aviation meteorology.