Sistema Regional de Información
en línea para Revistas Científicas de América Latina,
el Caribe, España y Portugal

Herbal drugs are widely used for medicinal purposes, but their quality can be compromised by adulterations and deficiencies in control processes. In Uruguay, botanical analyses of these products remain scarce despite their growing demand. This study evaluated the botanical quality of 42 herbal drug samples used by local companies and laboratories prior to commercialization. The samples were analyzed through organoleptic characterization, macroscopic and microscopic examination, following the criteria established in Decree 289/018 of the Uruguayan Ministry of Public Health. It was determined that 69 % of the samples corresponded to materials of acceptable quality.The main causes of non-compliance were the presence of plant organs that are not part of the herbal drug, the presence of fungi, insects or inorganic material, and the use of incorrect species. In addition, cases of adulteration involving botanically related species were analyzed in detail, particularly in drugs from the families Apiaceae, Malvaceae and Asteraceae, selected for their recurrence in commercial samples evaluated in previous years. The results highlight the need to strengthen botanical quality controls in themarketing of herbal drugs in the country to ensure their authenticity, efficacy, and safety.

Neural Networks have significantly evolved, particularly in their application to computer vision. This paper presents a comprehensive comparison of different versions of YOLOv8, such as YOLOv8n, YOLOv8s, YOLOv8m, YOLOv8l, and YOLOv8x for the detection of pests in bean plants, leveraging the capabilities of Convolutional Neural Networks. To train the neural network using different versions of YOLOv8, identical conditions were applied, such as the amount of environment light, the number of labeled images, epochs, and batch size. The results indicate that, as the complexity of the YOLO model increases, the training time escalates significantly. This increase corresponds to a more detailed data processing approach in advanced models. The results also provide insight on which model emerges as the most balanced option, offering the highest precision without compromising too much on speed.  One of the models achieves the highest precision,  making it reliable for accurate object detection but the speed is slow compared with other models. Otherwise exceptional precision makes it ideal for tasks where accurate identification is critical. The slight reduction in speed does not significantly hinder its overall performance in contexts where precision and detection distance are prioritized.

In this study, a modular avionics system solution is presented for fixed-wing fully autonomous controlled swarm unmanned aerial vehicles that can successfully detect within-visual and beyond-visual targets in the mission area as many times as possible thanks to deep learning, lock on to the target for the desired period of time and destroy it, evade the locks of other armed unmanned aerial vehicles with maneuvers and continuously transfer the information it receives to ground stations. Armed unmanned aerial vehicles provide reconnaissance, surveillance and intelligence opportunities as well as detection and confirmation of within-visual and beyond-visual targets in risky, complex conflict zones. Avionics system architecture consists of three units: the air platform consisting of swarm unmanned aerial vehicles, the ground station where swarm unmanned aerial vehicles are tracked throughout the mission flight and the IoT platform. Swarm armed unmanned aerial vehicles can take off autonomously and calculate the orientation and direction information of all aircraft in the airspace. Accordingly, the locking process is performed via the high-resolution/wide-angle imaging system by focusing on the appropriate aircraft that can be tracked. With the artificial intelligence (AI) algorithms to be developed for the continuity of the locking process, it is ensured that the swarm armed unmanned aerial vehicles can perform appropriate maneuvers. Aircraft equipped with appropriate systems for autonomous flight have a mission computer to run the algorithms related to the task. A joint study is carried out with the ground station and IoT platform to ensure the controllability and tracking of the system during the tracking of the swarm armed unmanned aerial vehicles. In order for these specified design points to be successfully performed during the mission, the system has a stable and agile flight control.

Optimizing the allocation of gasoline stations to refineries and depots enhances service coverage and response times, particularly during contingencies. Integrating biofuel production facilities supports sustainable energy initiatives and fuel supply. This study formulates an approach to optimize the allocation of gasoline stations to existing production and storage facilities while planning for bio-diesel plant expansion. It considers geographic distribution, balanced load allocation, minimal distances, and expansion costs. Using accurate data from Mexico, an extended k-means algorithm was developed to address the location/allocation problem involving gasoline stations, refineries, depots, and future biofuel facilities. Results demonstrate that additional biofuel facilities can alleviate service loads on existing infrastructure while enhancing sustainability and reducing environmental impact. These findings provide valuable insights for policymakers and urban planners aiming to improve fuel distribution networks and promote sustainable energy solutions.

This article presents an analysis of computer vision algorithms for Lane Maintenance Assistants (LMA), comparing traditional  methods with Convolutional Neural Networks (CNNs). The objective is to evaluate their effectiveness under diverse driving conditions using recognized databases and testing in both real and simulated environments. A proprietary database containing scenarios from the state of Morelos was also used. Experiments covered adverse conditions, such as rain (light, moderate, heavy), solar glare, road shadows, curves, and night driving with/without artificial lighting. Fog simulations included uniform,  heterogeneous, cloudy, and combined types. Results showed traditional methods perform well in normal conditions but struggle in complex scenarios like heavy rain, sharp curves, and poor lighting. CNN-based algorithms like SCNN and VGG16 demonstrated greater adaptability and accuracy in challenging environments, outperforming traditional methods. This study highlights the advantages of deep learning in improving road safety under adverse conditions.

Fashion themes, along with suitable colours and design styles, are crucial for creating garments that align with consumer preferences. Traditionally, designers have relied on intuition; however, fuzzy logic has gained recognition as a formal method for modelling the ambiguous nature of fashion design. By applying fuzzy techniques, designers can quantify the relevance of colours and styles to specific fashion themes, leading to a more structured decision-making process. This study aims to develop a fuzzy modelling framework for women’s swimwear design, formalising the relationships between fashion themes, colours, and styles. Expert evaluations contributed to establishing relational matrices for colour elements and design styles, which were validated by comparing them with expert opinions. The ultimate goal of this research is to develop practical guidelines for designers to enhance the alignment of fashion themes with colours and styles, particularly for mass-customised swimwear, utilising a data-driven approach.

Chemists often rely on computational prediction techniques to optimise laboratory experiments. However, selecting and configuring these techniques can be a daunting task for those without a background in computer science. A potential solution arises when chemists collaborate with computer scientists to define and implement the prediction task. Based on this collaboration, we propose a novel architecture that reflects key aspects of this interaction: the expertise of computer scientists, their commitment to addressing the prediction task, and the discussions that facilitate consensus. Our proposed architecture enables chemists to generate solution proposals across the various stages of the prediction process, leveraging consensus-based decisions made by intelligent agents and incorporating a lazy learning approach.

Violence against women is one of the most common human rights violations, with its most extreme form being femicide. In this context, we considered it relevant to demonstrate how artificial intelligence tools and geospatial analysis techniques can contribute to a better and faster analysis of these crimes. In this study, we analysed femicides that occurred in Mexico between 2014 and 2022. Our data source comprised digital news articles from leading Mexican newspapers. The study begins with the preprocessing of texts and the detection of those mentioning femicide. Subsequently, using unsupervised learning models, we grouped the texts according to their semantic similarity. We then employed deep learning models to classify each crime according to its specific characteristics. Finally, we used spatial analysis tools to detect geographic patterns in the occurrence of these crimes in the metropolitan area of the Valley of Mexico, analysing the automatically detected characteristics as variables.

Space exploration rovers, one of the most important tools in today's exploration studies, are designed to obtain scientific data by examining the surfaces of different planets and to seek answers to questions of great importance to humanity. In this study, a controller with nonlinear governing equations and path planning are carried out on advanced avionic systems for difficult conditions in order to autonomously select the shortest path at a desired trajectory and speed for a differentially driven space exploration rover. While the rover leaves its traces on the difficult surfaces of different planets, it undertakes many important tasks from the analysis of geological formations to the potential traces of life. Differential drive allows these rovers to move flawlessly even on rough surfaces without making sharp turns. It is observed that the RRT algorithm can be used in the implementation of path planning, unlike other path planning algorithms, to reach the target in a shorter time. This opens the door to new scientific understandings by carrying space exploration to previously inaccessible areas.

This paper introduces a model that detects harassment cases across different social media platforms using natural language processing and deep learning. Therefore, this model forms part of a system that can identify this problem in various texts.