The Congo Basin is an ecological and socio-economic linchpin facing escalating threats from floods, droughts, and soil erosion. Its agricultural sector, dominated by smallholder farmers, is particularly vulnerable. This vulnerability is exacerbated by a critical data deficit, with obsolete maps and sparse ground stations creating an informational void that renders traditional, reactive monitoring methods inadequate.

This paper argues that integrating Geographic Information Systems (GIS) with Artificial Intelligence (AI) and satellite data represents a paradigm shift for proactive natural risk assessment. We demonstrate how machine learning (ML) and deep learning automate geospatial analysis. Synthetic Aperture Radar (SAR) from Sentinel-1 penetrates persistent cloud cover to monitor floods and ground movement, while optical data from Sentinel-2, processed by convolutional neural networks (CNNs), enables high-frequency detection of deforestation, crop stress, and erosion. Crucially, AI's power extends to predictive modeling; by analyzing historical and real-time climate, topographic, and land-use data, ML models can forecast extreme events, shifting risk management from reactive to anticipatory.

To operationalize this potential, we conceptualize a prototype decision-support tool. This platform synthesizes multi-source satellite data with predictive AI models to generate dynamic, real-time risk zone maps. The demonstration showcases how the tool leverages spectral indices like NDVI (vegetation health) and NDWI (soil moisture) to produce intuitive visualizations. These maps pinpoint emerging threats—such as imminent flood zones or areas of severe soil nutrient loss—with surgical precision. This AI-driven cartography empowers diverse stakeholders: agricultural agencies can target aid, environmental bodies can combat illegal logging, and urban planners can enforce safe zoning. Ultimately, it provides farmers and local decision-makers with localized, predictive insights to safeguard livelihoods, optimize resources, and build foundational resilience against climate-induced disasters.

Light Detection and Ranging (LiDAR) has become a powerful technology for acquiring dense 3D point clouds with high spatial accuracy, enabling volumetric analysis and topographic characterization at multiple scales. Terrestrial Laser Scanner (TLS), here after LiDAR, have expanded their use from natural resource management to the structural health monitoring (SHM) of critical civil infrastructure in dynamic and hazard-prone environments. This study focused on inspecting and documenting infrastructure conditions along the three main rivers of the Monterrey Metropolitan Area: Pesquería, La Silla, and Santa Catarina. The objective was to assess structural integrity and identify elements requiring preventive maintenance or removal. Photographic datasets were compared with Google Street View and Google Earth Pro imagery to geolocate and detect visible changes. Based on this initial analysis, 5 to 6 structures per river were selected for detailed surveys using TLS equipment. TLS data were processed to create detailed plans and 3D models of each structure. Technical files were completed to document structural characteristics, condition assessments, repair proposals, and risk and accessibility analyses. This study demonstrates an effective methodology for SHM of riverine infrastructure. It also emphasizes the need for multidisciplinary expertise—including engineering, geospatial analysis, and urban planning—to support evidence-based public policy. The resulting datasets can inform hydrological and hydraulic modeling, enhance resilience assessments, and guide urban planning, construction strategies, and civil protection efforts in one of Mexico’s most vulnerable metropolitan regions.

Durante 2025 se desarrolló una investigación orientada a caracterizar la distribución, estructura y condición de los magueyes mezcaleros en el ejido Marcela, una comunidad conformada por 45 habitantes y situada a 2,496 m de altitud en el municipio de Miquihuana, al sureste del estado de Tamaulipas. Este ejido alberga un ecosistema de montaña donde predominan Agave montana y Agave gentryi, especies silvestres endémicas del noreste de México, de alto valor ecológico, cultural y económico, consideradas base para la producción de mezcal artesanal de alta calidad.

Por consiguiente, el objetivo del levantamiento fue caracterizar espacialmente las especies de Agave identificadas, evaluando su distribución, cobertura y morfometría mediante el procesamiento integrado de nubes de puntos LiDAR (aéreo – terrestre) y ortomosaicos generados con UAV Matrice 350 RTK. Este primer acercamiento se enmarca en una estrategia que articula el uso de tecnologías avanzadas con metodologías tradicionales de campo aplicadas a la hidráulica y el monitoreo de infiltración del agua en el suelo. Para ello se conformaron equipos interdisciplinarios en geomática, ecología, ingeniería ambiental e hidráulica.

Aunado a lo anterior, La densidad promedio de la nube de puntos alcanzó 342 puntos por metro cuadrado, con una proporción de celdas no conformes menor al 2%. Por parte de los ortomosaicos se obtuvo una resolución espacial fue de 3.01 cm/píxel. El análisis preliminar de la nube LiDAR permitió identificar una densidad promedio de 4 ± 1 individuos de agave por pino, reflejando patrones espaciales ligados a la cobertura arbórea y las condiciones topográficas. La información generada constituye una base sólida para estimar la densidad y expansión de las poblaciones naturales, además de sustentar el diseño de estrategias de conservación y aprovechamiento sustentable del recurso mezcalero, alineadas con el desarrollo comunitario y la gestión ambiental responsable.

Las empresas a nivel mundial enfrentan un desafío significativo en la selección de proveedores dentro de sus operaciones de transporte internacional, debido a altos niveles de incertidumbre que afectan la efectividad de respuesta de los múltiples actores logísticos. Este artículo presenta un sistema predictivo basado en Deep Learning orientado a la evaluación y selección sostenible de proveedores mediante la integración de información logística y geoespacial. El modelo utiliza datos históricos de desempeño, tiempo de tránsito y confiabilidad, aplicando redes neuronales recurrentes (LSTM y GRU) entrenadas con información proveniente de bases SQL empresariales y plataformas de seguimiento de carga (AIS). La metodología CRISP-DM guía el proceso desde la recolección de datos hasta la validación y despliegue, culminando en un prototipo funcional desarrollado en Python. Los resultados muestran una mejora significativa en la precisión de predicción del cumplimiento de proveedores, fortaleciendo la toma de decisiones estratégicas en transporte marítimo internacional.

During World War II, Ecuador authorized the construction of two U.S. military bases as part of a strategic defense agreement. These bases, located on South Seymour Island (Baltra) and Punta Carnero in mainland Ecuador, played a key role in protecting the Panama Canal and monitoring the eastern Pacific. In early 1942, U.S. Navy Seabees built the Baltra airbase, which included two airstrips, over 200 buildings, and housed more than 3,000 people. The base significantly altered the island’s landscape and ecosystem. In 1946, the U.S. returned both bases to Ecuador, but many facilities were dismantled, abandoned, or destroyed.

This project proposes a 3D digital reconstruction of the Baltra base as it existed during the war, using photogrammetry and remote sensing. We applied a multidisciplinary approach combining historical research, remote sensing, and modern surveying. Key steps included researching, analyzing, digitizing and georeferencing historical aerial photographs and maps, reviewing archival documents for physical descriptions of the base, and processing remote sensing data. A current topographic survey using drone imagery will generate digital elevation models (DEMs). Archival research was conducted in Ecuador (INOCAR, IGM, and the Ministry of Foreign Affairs) and the U.S. (NARA). Satellite imagery was sourced from Planet and readily available sources as Google and Microsoft. Additionally, we used visual interpretation to create a 3D model of relevant structures of the base. Digital archaeology aims to preserve the materiality and memory of at-risk sites. By reconstructing Baltra, we can visualize historical spatial layouts and contribute to research, education, and conservation.

Baltra’s human history has received little scholarly attention compared to earlier settlements in the archipelago. While some barracks were reused in nearby towns, most of the site has been looted or left to deteriorate. The digital model helped in assess the site's current condition, highlighted conservation priorities, and documented human impact. Though short-lived, the U.S. presence left a lasting material, ecological, and social legacy.