Conference Agenda

Window cleaning in high-rise buildings is a high-risk activity that demands automated solutions to ensure industrial safety and operational efficiency. This paper presents the design, prototyping, and validation of a modular mobile robot with active suction, optimized for glass facade maintenance. The system utilizes high power brushless motors to generate controlled suction, paired with a differential locomotion system for stable vertical displacement. A key technical contribution is its reconfigurable architecture, allowing the integration of interchangeable modules for cleaning, structural inspection, or monitoring. The firmware, developed using Object-Oriented Programming (OOP), enables robust logical management of peripherals and sensors, facilitating seamless system configuration. To validate the approach, a functional prototype was evaluated under real conditions, focusing on load capacity, aerodynamic stability, and adhesion reliability. Results demonstrate constant suction and effective trajectory control, significantly reducing human risk in critical environments. Furthermore, this research establishes the basis for future autonomous navigation and sensor integration in large scale urban settings. Consequently, this mechatronic development represents a substantial advancement in automating specialized services for smart city infrastructure. By providing a versatile and safe platform, this work addresses the inherent challenges of high-altitude maintenance while establishing a scalable framework for diverse vertical robotic applications in modern construction.

This paper presents the design and implementation of an autonomous environmental monitoring platform based on an unmanned aerial vehicle equipped with a Pixhawk 6C flight controller and a dual ESP32 architecture. The onboard ESP32 functions as a dedicated sensing module that interfaces with MQ-2, MQ-9, and MQ-135 gas sensors to estimate pollutant levels associated with combustion and urban activity. A second ESP32 located at the ground station records the data received via a Wi-Fi link. Processed measurements are simultaneously transmitted to the Pixhawk through custom MAVLink messages, providing a redundant communication channel when Wi-Fi coverage degrades. The integrated system supports automated waypoint missions, modular power distribution, and real-time telemetry, enabling repeatable data-collection flights over predefined survey areas without manual piloting. Field tests were conducted in Caguas and Juncos, Puerto Rico, to compare environmental conditions in two urban regions with differing levels of population density and vehicular activity. Results demonstrate that the platform provides a scalable low-cost method for geo-referenced environmental monitoring using open-source hardware suitable for multi-vehicle deployments.

This paper presents the development of a low-cost electronic system designed to indirectly monitor the organoleptic attributes of coffee during the post-harvest stage, with the purpose of differentiating states such as cherry coffee, honey coffee, and dry parchment coffee. The system integrates MQ3 and MQ135 gas sensors, used to detect compounds associated with fermentation processes and the release of volatile organic compounds, together with a DHT11 sensor for measuring ambient temperature and humidity. Data acquisition and processing are carried out through an ESP32 microcontroller programmed in the Arduino® IDE, enabling local storage on a microSD card, visualization through a graphical interface developed in MATLAB®, and wireless transmission to the IoT platform ThingSpeak™ for remote monitoring. Experimental tests conducted in different post-harvest states revealed differentiated patterns in relative alcohol concentration, levels of volatile compounds, and humidity, consistent with the evolution of fermentation and drying processes. Qualitative validation with a coffee expert confirmed the potential of the system as an accessible and useful tool for post-harvest analysis and monitoring, contributing technological innovation to quality control in the coffee sector.

This project presents the design, implementation, and validation of an autonomous service robot intended to guide visitors, students, and staff within university campuses. The initiative addresses the architectural complexity of the campus and aims to optimize travel times while reinforcing the institution’s technological innovation.

The system is built on a hierarchical control architecture: an ESP32 microcontroller acts as the master unit and web server, managing the user interface and high‑level logic, while a secondary unit performs closed‑loop control of the actuators. Mechanically, the robot uses a Direct Drive system with high‑torque wiper motors and A36 steel shafts in a cantilever configuration, validated through Von Mises stress analysis to support loads up to 20 kg with a safety factor greater than 3.0.

Navigation is achieved through a hybrid trajectory recording and playback algorithm. A QMC5883L magnetometer with PID control ensures accurate heading and rectilinear motion, while a TFmini‑S LiDAR sensor halts the robot when dynamic obstacles are detected within 1.5 m. A NEO‑M8N GPS module provides checkpoint validation through geolocation.

Human–Machine Interaction is carried out through a Progressive Web Interface stored in the robot’s memory and accessible via its Wi‑Fi Access Point mode, complemented by text‑to‑speech audio feedback. Experimental results demonstrate stable navigation, effective obstacle detection, and intuitive interaction, meeting safety and functionality requirements for academic environments.

This article presents the development of a vision module for robotic navigation based on people tracking, using ROS2 for speed control via a publisher node embedded in a NVIDIA-Jetson Nano board card and a subscriber node with an ESP32 microcontroller. This integrates the speed control of the mobile device according to the detection distance of a person using a YOLO deep network, to maintain a safe distance between the robot and the user. The training characteristics of the latest YOLO architecture based on version 12 are presented, and PID control is used with a Pololu DC motor with an encoder integrated into the subscriber node, successfully achieving distance control based on the user's proximity to the robot.

This work aims to describe the development and implementation of a remote monitoring system for environmental variables in greenhouse crops using Internet of Things (IoT) technologies. A four-layer architecture was designed, composed of four sensing nodes based on the ESP32 microcontroller. These nodes can measure relative humidity, ambient temperature, and light intensity, and wirelessly transmit these variables to a central gateway implemented on a Raspberry Pi 4. The gateway then forwards the data to a web platform for visualization. Ambient temperature was observed to range between 13 °C and 29 °C, relative humidity remained above 50%, and solar radiation during peak hours ranged from 4,000 to 12,000 lux. The system proved to be a viable and easily implemented solution for Agriculture 4.0, providing valuable information for making decisions regarding the most favorable growing conditions.