The purpose of this project is to develop a tool capable of performing a comprehensive analysis of people flow at the Escuela Superior de Cómputo (ESCOM), combining both descriptive and predictive approaches. To achieve this, an autonomous UAV is deployed to fly over the campus, capturing a continuous video stream at different times of the day. Equipped with a thermal camera and deep learning models, the system performs real-time person detection and counting. The collected information is processed, organized, and stored locally, enabling the application of statistical methods and machine learning techniques to characterize behavior patterns and identify critical areas of congestion. Beyond descriptive analysis, the project incorporates predictive modeling to anticipate crowd dynamics and support decision-making. Using the Prophet algorithm, time-series data is decomposed into trend, seasonality, and residual components, producing reliable forecasts of people flow across specific zones. These results are integrated into the ESCOM-Flow application, which centralizes drone mission control, real-time monitoring, and automated reporting. In this way, the system not only enhances crowd management at ESCOM but also provides a scalable framework for future applications in other high-demand environments.