Automated Guided Vehicles (AGVs) represent a transformative advancement in the automation of transport operations, facilitating unmanned mobility within a wide array of environments, including production lines, warehouses, freight hubs, and terminal operations. In container terminals, where AGVs are increasingly deployed, the routing of these vehicles is a critical task aimed at minimising operational inefficiencies such as travel time, fuel consumption, and overall transportation costs. Routing in this context refers to the determination of optimal paths for a fleet of AGVs, which must satisfy a variety of operational constraints while also adhering to predefined user requirements. Given the high complexity of these problems, characterised by a large solution space, finding exact solutions is computationally intractable for most scenarios. As a result, heuristic methods are commonly employed to approximate optimal solutions. Among the various heuristic techniques, the nearest neighbor algorithm and Tabu search have been identified as promising approaches for determining efficient AGV routes in container terminal environments. These methods are applied to identify paths that minimise travel distance and time, enhancing resource utilisation and improving the overall reliability of goods delivery. The application of these algorithms is expected to lead to a significant reduction in the number of kilometres travelled by AGVs, thereby lowering operational costs and improving service efficiency. This paper examines the efficacy of the "nearest neighbor" and Tabu search algorithms in the context of AGV routing at container terminals, highlighting their potential to optimise fleet operations in the face of complex logistical challenges. Emphasis is placed on the comparative analysis of both algorithms, with a focus on their ability to approximate optimal solutions in dynamic and highly constrained environments.

An innovative context-aware fuzzy logic transmission map adjustment method is proposed for road image defogging, aimed at improving visibility and clarity under varying fog conditions. Unlike conventional defogging techniques that rely on a uniform transmission map, the presented approach introduces a fuzzy logic framework that dynamically adjusts the transmission map based on local fog density and contextual factors. Fuzzy membership functions are employed to classify fog density into low, medium, and high categories, enabling an adaptive and context-sensitive adjustment process. Road images are segmented into distinct regions using edge detection and texture analysis, with each region treated independently to preserve critical details such as road markings, lane boundaries, and traffic signs. A key contribution is the integration of proximity-based adjustments for areas near high-intensity light sources, such as streetlights, to maintain brightness and enhance visibility in illuminated zones. The final transmission map is generated through the combination of fuzzy density-based adjustments and an iterative Gaussian filter, which smooths transitions and minimizes potential artifacts. This approach prevents over-darkening while enhancing contrast, even in dense fog conditions. Experimental results demonstrate that the proposed method significantly outperforms traditional defogging techniques in terms of brightness, contrast, and detail retention. The results underscore the utility of fuzzy logic in road image defogging, offering a robust solution for applications in autonomous driving, surveillance, and remote sensing. This method sets a new benchmark for visibility enhancement in challenging environments, providing a high-quality, adaptive solution for real-world applications.

Advanced ship detection technologies play a critical role in improving maritime safety by enabling the rapid identification of vessels and other maritime targets, thereby mitigating the risk of collisions and optimizing traffic efficiency. Traditional detection methods often demonstrate high sensitivity to minor variations in target appearance but face significant limitations in generalization, making them ill-suited to the complex and dynamic nature of maritime environments. To address these challenges, an enhanced ship detection method, referred to as YOLOv5-SE, has been proposed, which builds upon the YOLOv5 framework. This approach incorporates attention mechanisms within the backbone network to improve the model's focus on key features of small targets, dynamically adjusting the importance of each channel to boost representational capacity and detection accuracy. In addition, a refined version of the Complete Intersection over Union (CIoU) loss function has been introduced to optimize the loss associated with target bounding box prediction, thereby improving localization accuracy and ensuring more precise alignment between predicted and ground-truth boxes. Furthermore, the conventional coupled detection head in YOLOv5 is replaced by a Decoupled Head, facilitating better adaptability to various target shapes and accelerating model convergence. Experimental results demonstrate that these modifications significantly enhance ship detection performance, with mean Average Precision (mAP) at IoU 0.5 reaching 94.9% and 95.1%, representing improvements of 3.1% and 1.2% over the baseline YOLOv5 model, respectively. These advancements underscore the efficacy of the proposed methodology in improving detection accuracy and robustness in challenging maritime settings.

Crowd logistics (CL) represents an innovative model within the logistics sector, leveraging the participation of individuals to enhance service provision, optimize resource utilization, and reduce operational costs. Among the various applications of CL, crowd distribution has emerged as one of the most prevalent methods. This study introduces a Multi-Criteria Decision-Making (MCDM) framework for the selection of CL platforms, examining key factors that contribute to their success. A comprehensive review of relevant literature and an in-depth analysis of both domestic and global platforms were conducted, revealing critical performance indicators for successful platform implementation. The Step-wise Weight Assessment Ratio Analysis (SWARA) and Measurement of Alternatives and Ranking according to Compromise Solution (MARCOS) methods were employed to evaluate essential criteria, including cost efficiency, delivery speed, reliability, environmental sustainability, flexibility, and customer support quality. The results of this analysis demonstrate that platforms such as Company 1, Company 2, and Company 3 have achieved market dominance in Serbia, attributed to their optimal balance across these performance criteria. This study’s proposed model serves as a practical tool for businesses and consumers seeking to select the most suitable CL platforms, while also providing actionable insights for further enhancement of logistics systems. The findings contribute to the growing body of knowledge on CL, highlighting the importance of comprehensive evaluation in the selection process.

Foggy road conditions present substantial challenges to road monitoring and autonomous driving systems, as existing defogging techniques often fail to accurately recover structural details, manage dense fog, and mitigate artifacts. In response, a novel defogging model is proposed, incorporating Pythagorean fuzzy aggregation, Gaussian Mixture Models (GMM), and the level-set method, aimed at overcoming these limitations. Unlike conventional methods that depend on fixed priors or oversimplified haze models, the proposed framework leverages the advantages of Pythagorean fuzzy aggregation to enhance contrast and detail restoration, GMM to estimate fog density robustly, and the level-set method for precise edge preservation. The performance of the model is quantitatively assessed, revealing a Peak Signal-to-Noise Ratio (PSNR) of up to 37.1 dB and a Structural Similarity Index (SSIM) of 0.96, which significantly outperforms existing defogging techniques. Statistical analyses further confirm the robustness of the approach, with a p-value of less than 0.001 for key performance metrics. Additionally, the model demonstrates an execution time of 0.07 seconds, indicating its suitability for real-time road monitoring applications. Qualitative assessments highlight the model's ability to restore natural road colours and maintain high structural fidelity, even under conditions of dense fog. This work provides a promising advancement over current methods, with potential applications in autonomous driving, traffic surveillance, and smart transportation systems.