The early and precise identification of the different phenological stages of the bean (Phaseolus vulgaris L.) allows for the determination of critical and timely moments for the implementation of certain agricultural activities that contribute in a significant manner to the output and quality of the harvest, as well as the necessary actions to prevent and control possible damage caused by plagues and diseases. Overall, the standard procedure for phenological identification is conducted by the farmer. This can lead to the possibility of overlooking important findings during the phenological development of the plant, which could result in the appearance of plagues and diseases. In recent years, deep learning (DL) methods have been used to analyze crop behavior and minimize risk in agricultural decision making. One of the most used DL methods in image processing is the convolutional neural network (CNN) due to its high capacity for learning relevant features and recognizing objects in images. In this article, a transfer learning approach and a data augmentation method were applied. A station equipped with RGB cameras was used to gather data from images during the complete phenological cycle of the bean. The information gathered was used to create a set of data to evaluate the performance of each of the four proposed network models: AlexNet, VGG19, SqueezeNet, and GoogleNet. The metrics used were accuracy, precision, sensitivity, specificity, and F1-Score. The results of the best architecture obtained in the validation were those of GoogleNet, which obtained 96.71% accuracy, 96.81% precision, 95.77% sensitivity, 98.73% specificity, and 96.25% F1-Score.

In this work, DonkieTown is introduced, an affordable and scalable platform for research on autonomous vehicles. The experimental framework was developed in the Robot Operative System (ROS). The platform integrates multiple small scale autonomous vehicles called Asinus Cars, which are equipped with at least a camera, odometer, and onboard computer. The vehicles are Differential Drive Robots (DDR), forced by software to behave as car-like vehicles. DonkieTown incorporates a low-cost localization system to provide the real-time vehicles' pose, by means of external cameras which detect ArUco markers, then Kalman Filters (KF) are used to track and estimate the pose of each vehicle. The platform includes a base station computer with a graphical interface for monitoring the system. DonkieTown also includes a series of algorithms to facilitate autonomous driving, such as communication, tracking, object detection, obstacle avoidance, control, trajectory tracking, etc. Moreover, a centralized vehicular network is implemented to allow communication between the agents and the base station, where the agents can share information about their state, obstacles, maneuver intentions, etc. To facilitate the research on autonomous cars in Latin America, the developed libraries are released as open source. Real-time experiments demonstrate the performance of DonkieTown in autonomous driving missions, such as following a lane while avoiding Donkey-like obstacles, and collaborative autonomous driving in convoy.

This document presents the design of a digital PID control for a Stewart-Gough platform, delimited by six degrees of freedom (DoF) that allow the final effector to have displacement in the XYZ axes and rotation with warpage, pitch, and yaw restrictions. It includes the study and resolution of the direct and inverse kinematics of the platform, as well as the workspace described by the final effector and its corresponding simulation of movements and joints to study extreme points and possible singularities. From the definition of characteristics, the CAD design generated from the generalized mathematical model of the public domain, and the general selection of materials for the construction of the functional prototype, a study of applied forces is generated to observe the points with stress concentrators, the safety factor, and possible deformations. The estimation of the sampling period for the selection of the microcontroller and an approximate definition of the response time are also considered. The development of this prototype and its documentation are proposed as didactic material for the study, design, and control of parallel mechanisms.

Today, tilapia farming is a standardized and well-defined process, however, fish farmers do not have the necessary tools to ensure that the process is being carried out properly in terms of portions, especially when it comes to large ponds, feeding the school of fish in them becomes a task that requires time and effort. In this work, a project proposal was developed to help fish farmers with the design and construction of a prototype aquatic mobile automatic feeder that, as its name says, allows feeding the fish by moving around the pond and with the correct ratio. In addition, the design of a human-machine interface was developed so that the operator could control the prototype, including the ability to decide when to release the feed. This project is intended to help facilitate the fish farmer's task of feeding the school of fish using the mechatronic design methodology.

Lane keeping under degraded visual confidence remains challenging because most existing approaches focus either on improving lane-feature extraction or on estimating vehicle motion, while giving less attention to how unreliable visual measurements should modify the estimator–controller interaction in a physical closed-loop system. This paper presents a confidence-gated Kalman lane-keeping architecture for a 1:10-scale autonomous vehicle. The methodological contribution lies in the direct coupling of visual confidence, state estimation, and steering control: unreliable lane measurements are down-weighted through confidence-dependent measurement noise, while the propagated lane-relative state remains available to the controller. The primary experimental unit is the run, defined as one logged lap under one control configuration; frame-level summaries are retained only as historical reproducibility material. In the run-level comparison, the autonomous vision, inertial measurement unit (IMU)-feedforward, and Kalman filter (KF) group KF_G1—the first inferential KF generation—had lower mean absolute error than the human baseline, while vision and KF_G1 had overlapping run-level confidence intervals for absolute error. KF_G1 shifted the mean signed bias closer to the lane reference than the vision and IMU groups, but with higher run-level spread than vision. KF_G2, the second observational KF generation, is reported only as an observational generation comparison, so no causal claim is made for its estimator–controller correction weight $K_s=0.15$ setting. KF_G3, the single descriptive adverse-condition KF run, is reported without an estimable confidence interval or population-level adverse-illumination inference. A further limitation is that the inertial prediction pathway was inactive in the logged Kalman-filter runs because the inertial coupling coefficient $\alpha =0$. The main contribution is a reproducible run-level evaluation of confidence-gated estimator–controller coupling that distinguishes supported evidence from observational and descriptive cases.