Conference Agenda

This article presents the development of an innovative tool based on a geolocation-enabled web application designed to monitor and report domestic animals in vulnerable conditions. The primary objective is to centralize and validate citizen-submitted reports, thereby reducing response times and strengthening the operational capacity of rescuers, organizations, and relevant authorities. The methodology focused on building a functional prototype through the integration of state-of-the-art technologies, including a convolutional neural network (CNN) trained to automatically identify two animal categories—dogs and cats—along with interactive map-based visualization modules and database systems for report management.
The CNN achieved a peak validation accuracy of 98.75% and a final training accuracy of 98.55% for animal-category classification. In addition, usability and performance evaluations indicated stable operation and a favorable user experience.
Field tests were conducted in the province of Veraguas, Panama, to validate the prototype. The project represents a meaningful contribution to the adoption of technological solutions for animal protection, showing high user acceptance and strong potential for national-scale deployment, supporting community-based surveillance and the planning of more efficient interventions.