La detección del estro es clave para optimizar la concepción en vacas y la eficiencia reproductiva del ganado. El método convencional requiere observación continua, demandando mano de obra y tiempo. Desarrollamos un sistema basado en IoT que automatiza el monitoreo del estro mediante un dispositivo multisensor montado en el cuello de la vaca. Este recopila datos y los transmite vía LoRaWAN a un Gateway, que los envía a The Things Stack y luego a TagoIO para visualización y almacenamiento. En pruebas de campo, tras sincronizar el estro de una vaca en la Amazonía peruana, se recolectaron datos y se analizó su variación. El sistema registró información fisiológica y de comportamiento, evidenciando que en 72 horas aumentaron el desplazamiento y la temperatura corporal, indicando el estro.

El rendimiento académico es crucial para la gestión educativa, y la predicción de este puede optimizar la toma de decisiones. Este estudio tuvo como objetivo predecir el rendimiento académico de estudiantes universitarios mediante tres algoritmos de aprendizaje supervisado: K-Vecinos más cercanos (KNN), Naive Bayes (NB) y Árbol de Decisión (AD). Se utilizaron datos de 813 estudiantes, incluyendo variables socioeconómicas y académicas. Los datos fueron preprocesados y evaluados a través de métricas como precisión, recuperación, exactitud y AUC-ROC. El modelo KNN fue el más efectivo, alcanzando una exactitud del 81.97%, lo que lo convierte en la mejor opción para predecir el rendimiento académico. Aunque mostró una recuperación moderada, su precisión fue la más alta, demostrando un buen equilibrio en la clasificación de los estudiantes. En conclusión, KNN es una herramienta prometedora para que las instituciones educativas identifiquen estudiantes en riesgo y mejoren sus estrategias de intervención, ayudando a elevar el rendimiento académico y reducir la deserción.

Liquefied petroleum gas (LPG), composed mainly of propane and butane, is widely used as an energy source in residential, commercial, and industrial sectors; however, its high flammability poses a critical risk in the event of accidental leaks. In Peru, where LPG constitutes the main domestic energy source, leakage emergencies affect thousands of households each year. This pattern is replicated in developing countries with limited energy infrastructure. Early quantitative detection of propane, the predominant component of Peruvian LPG (~60%), is essential to prevent explosions, poisoning, and greenhouse gas emissions that hinder climate change mitigation strategies. This study presents PropNet-R, a convolutional neural network (CNN) designed to estimate propane concentrations (ppm) from thermal images. A dataset of 3574 thermal images synchronized with concentration measurements was collected under controlled conditions. PropNet-R, composed of four progressive convolutional blocks, was compared with SqueezeNet, VGG19, and ResNet50, all fine-tuned for regression tasks. On the test set, PropNet-R achieved MSE = 0.240, R² = 0.614, MAE = 0.333, and Pearson’s r = 0.786, outperforming SqueezeNet (MSE = 0.374, R² = 0.397), VGG19 (MSE = 0.447, R² = 0.280), and ResNet50 (MSE = 0.474, R² = 0.236). These findings provide empirical evidence that task-specific CNN architectures outperform generic transfer learning models in thermal image-based regression. By enabling continuous and quantitative monitoring of gas leaks, PropNet-R enhances safety in industrial and urban environments, complementing conventional chemical sensors. The proposed model contributes to the development of sustainable infrastructure by reducing gas-related risks, promoting energy security, and strengthening resilient, safe, and environmentally responsible urban systems.

Aquaculture is carried out in controlled environments to guarantee the health, growth, and survival of aquatic species, so it is essential to monitor water quality parameters. We designed and implemented a system based on the Internet of Things for monitoring dissolved oxygen and temperature in Amazonian fish larviculture. We based the design of the architecture on four layers: data acquisition, communication, cloud server and application. We carried out the experimental tests with the data collected by the system during seven days of culture of larvae of the species Piaractus brachypomus (paco) in the aquaculture center “La Cachuela”. The values obtained for dissolved oxygen varied between 7.43 ± 0.04 and 7.83 ± 0.02 mg L-1; regarding the water temperature, it fluctuated between 19.59 ± 0.45 and 25.45 ± 0.11 °C. We concluded that the system developed for monitoring dissolved oxygen and water temperature displays data obtained in real time, in both cases the values are within the tolerable ranges for the cultivation of this species. La acuicultura se desarrolla en entornos controlados para garantizar la salud, crecimiento y supervivencia de las especies acuáticas, por lo que resulta clave monitorear los parámetros de calidad del agua. Diseñamos e implementamos un sistema basado en Internet de las Cosas para monitoreo del oxígeno disuelto y temperatura en larvicultura de peces amazónicos. Basamos el diseño de la arquitectura en cuatro capas: adquisición de datos, comunicación, servidor de nube y aplicación. Realizamos las pruebas experimentales con los datos recolectados por el sistema durante siete días de cultivo de larvas de la especie Piaractus brachypomus (paco) en el centro acuícola “La Cachuela”. Los valores obtenidos para el oxígeno disuelto variaron entre 7,43 ± 0,04 y 7,83 ± 0,02 mg L-1; en relación con la temperatura del agua esta fluctuó entre 19,59 ± 0,45 y 25,45 ± 0,11 °C. Concluimos que el sistema desarrollado para monitoreo de oxígeno disuelto y temperatura del agua visualiza datos obtenidos en tiempo real, en ambos casos los valores se encuentran en los rangos tolerables para el cultivo de esta especie.

Acute malnutrition affects 3.1% of children under five years of age in Amazonian communities in Peru, where limited access to health services constrains timely nutritional assessment. In this context, this study aimed to develop, implement, and evaluate NutriRadar, a mobile application for automated childhood nutritional classification based on the anthropometric standards of the World Health Organization (WHO). The application was developed using a waterfall software development methodology and implements the calculation of the Weight-for-Height Z-score (WHZ) from basic anthropometric variables (weight, height, age, and sex). NutriRadar was designed with offline functionality, deferred data synchronization, and compatibility with low-end mobile devices to support operational use in Amazonian settings. Field validation was conducted in two early childhood education institutions in Puerto Maldonado, Peru, and included anthropometric assessments of 75 children aged 3–4 years. The application demonstrated stable offline operation, response times suitable for clinical practice, and nutritional classification results equivalent to the WHO Anthro reference tool. NutriRadar represents a viable and reproducible digital automation solution for the operational application of a deterministic WHO anthropometric protocol, contributing to the reduction of operational errors and strengthening standardized nutritional assessment in resource-limited Amazonian contexts.

Water service payment delinquency limits the availability of resources required to sustain operations and hinders the prioritization of collection activities. This study aimed to evaluate machine learning models for estimating the payment default risk of residential customer bills at EMAPAB S.A. A total of 264,382 bills from 6,358 customers were analyzed using 51 predictors derived from historical billing and payment records. Random Forest, CatBoost, and a feedforward neural network were compared using Optuna optimization, stratified and grouped ten-fold cross-validation, and an independent temporal test. Random Forest achieved the best performance on the temporal test, with an accuracy of 0.9037, an F1-score of 0.6333, a ROC-AUC of 0.8681, and a PR-AUC of 0.6892. Its probabilities were subsequently calibrated using Platt Scaling and stratified into low-, medium-, and high-risk levels, accounting for 74.61%, 13.96%, and 11.44% of the bills, respectively. The results show that the proposed approach can identify bills with higher default risk and support the prioritization of the collection portfolio, although its operational impact requires prospective validation.