Marta Silvia Tabares

The export of Hass avocado (Persea americana Mill.) from Colombia requires accurate determination of harvest maturity, currently assessed through destructive dry matter (DM) measurements that are wasteful and limited in throughput. The objective of the article is to propose a low-cost, non-destructive approach to determine the maturity of the Hass avocado crop based on machine learning techniques. The approach consists of a low-cost, non-invasive bioimpedance spectroscopy system operating in the 1–10 kHz range, featuring a custom Analog Front End (AFE) and a tetrapolar surface probe to mitigate skin contact resistance, which collects data for predictive models of avocado maturity. To evaluate the quality of the approach, a longitudinal field study (n = 100) was conducted in a commercial orchard in Cundinamarca, Colombia, tracking complex impedance features—Magnitude, Phase Angle, Resistance, and Reactance—of tagged fruits over 8 weeks across four measurement timepoints. The predictive performance of a classical chemometric model (PLS-DA), non-linear classifiers (SVM, Random Forest), and a temporal Deep Learning (LSTM) architecture was compared using a Stratified Group K-Fold Cross-Validation scheme to prevent data leakage across fruits from the same tree. The 4-electrode configuration successfully isolated mesocarp impedance, identifying the 5–7.2 kHz band as the most sensitive to physiological maturation. In turn, the LSTM model achieved a mean accuracy of 92.0% and an AUC of 0.94, outperforming the other models by 4.0% in mean accuracy. The results demonstrate that modeling the temporal trajectory of impedance, rather than single-point measurements, improves harvest maturity classification in Hass avocados, providing a scalable, low-cost alternative to destructive testing.

Digital startups are driving a profound transformation that permeates large organizations, synergies that allow them to maintain their position in modern markets. To fully understand this cycle of change, it is imperative to consider the role of Application Programming Interfaces (APIs), which enable transformation through innovative business models. In finance, it is an undeniable fact that Fintechs are actively using APIs to drive digital business. However, large banks face new and challenging obstacles to generate new value in collaboration with these innovative digital partners, or to acquire new capabilities individually through the strategic implementation of APIs. In this context, a systematic mapping of the literature is presented with the aim of showing a broad and structured view on the current state of knowledge related to API governance in the financial sector framed in a research project for this sector and specifically for banking collections. From the study it was concluded that APIs are indispensable in today's ecosystems, they allow consumers, officials, third-party developers, Fintech, retailers and other components such as algorithms, data, processes and functionalities to communicate, therefore, businesses need to be redesigned to define rules that allow them to compete and expand their margins, without APIs it is not possible to embrace the digital world.

Proper waste separation is essential for recycling. However, it can be challenging to identify waste materials accurately, especially in real-world settings. In this study, a systematic literature review (SLR) was carried out to identify the physical enablers (sensors and computing devices), datasets, and machine learning (ML) algorithms used for waste identification in indirect separation systems. This review analyzed 55 studies, following the Kitchenham guidelines. The SLR identified three levels of autonomy in waste segregation systems: full, moderate, and low. Edge computing devices are the most widely used for data processing (9 of 17 studies). Five types of sensors are used for waste identification: inductive, capacitive, image-based, sound-based, and weight-based sensors. Visible-image-based sensors are the most common in the literature. Single classification is the most popular dataset type (65%), followed by bounding box detection (22.5%). Convolutional neural networks (CNNs) are the most commonly used ML technique for waste identification (24 out of 26 articles). One of the main conclusions is that waste identification faces challenges with real-world complexity, limited data in datasets, and a lack of detailed waste categorization. Future work in waste identification should focus on deployment and testing in non-controlled environments, expanding system functionalities, and exploring sensor fusion.

Avocado is one of the most commercialized and profitable fruits in the international market. Its cultivation and production are centered in countries characterized by tropical and subtropical climatic conditions, many of them with emerging economies. Moreover, the use of technology is key to agricultural production improvement strategies. Using avocado crop data to forecast the potential impacts of biotic and abiotic factors, combined with smart farming technologies, growers can apply measures during a single production phase to reduce the risks caused by pests and weather variations. Therefore, this paper aims to distinguish the most relevant variables related to agroclimatic and phytosanitary events in avocado crops, their incidence on production and risk management, as well as the emerging technologies used for the identification and analysis of pests and diseases in avocados. A scientific literature search was performed, and the first search found 608 studies, and once the screening process was applied, 37 papers were included in this review. In the results, three research questions were answered that described the pests and diseases with high impact on avocado production, along with the data sources and the principal enabling technologies used in the identification of agroclimatic and phytosanitary events in avocados. Some challenges and trends in the parameterization of the technology in field conditions for data collection are also highlighted.

This article presents a predictive model to determine possible obstructions in the sewerage system of the Valle de Aburrá metropolitan area in Colombia. This city has special characteristics for being located in a valley surrounded by high mountains, susceptible to rainfall almost all year round and with a population of over 4 million inhabitants. To achieve this, we apply the CRISP-DM methodology. Initially, the way how the system determines the obstructions through preventive and corrective maintenance was identified. Subsequently, the data provided by the main assets that make up the system, especially the spillways, were understood; then, the dynamic and exogenous variables that could influence the occurrence of obstructions were selected. A case study with a dataset of 3980 records with information between the years 2018 - 2021, facilitated the application of feature selection techniques to analyze the most influential variables, and the predictive model was developed to identify possible obstructions in the system. We use algorithms such as neural networks, decision trees, random forests, logistic regression, and support vector machines. The results of the neural networks yielded an accuracy of 84% that was calculated with the cross-validation method and an F1 Score of 70%. In the evaluation, the model showed that it was determined by the variables number of previous obstructions, density of afferent clients, previous maintenance, and the average rainfall of the last days. These results expand the decision-making capacity to define preventive maintenance and reduce corrective maintenance in the spillways that make up the sewerage system. Finally, it is concluded that the model can correctly predict the possible obstructions, however, it is necessary to deploy it in the complete system so that the prediction increases its level of accuracy.