Security in academic laboratories involves not only the protection of physical resources but also the control of access to environments where research and equipment are concentrated. Conventional systems based on keys or access cards often present limitations related to loss, duplication, or lack of traceability. This project presents the development of an access control system based on facial recognition, designed for use in research laboratories. The solution integrates an ESP32-CAM module for image capture and facial identification, an ESP32 microcontroller to manage access locks, and a TFT display to show visual feedback. A web interface was implemented to register users and record access events, allowing monitoring of entry times and user identification. The development followed a structured approach: initial research and planning, definition of requirements, component selection, and staged implementation. Functionalities such as image processing, lock control, and access logging were progressively integrated and tested. The system was evaluated in a controlled environment simulating real laboratory access scenarios. It correctly recognized registered users, denied entry to unregistered individuals, and recorded access data without communication failures or delays. All components functioned as expected, including wireless data transmission and interface synchronization. The proposed system provides an alternative to traditional key-based methods by offering traceability, ease of use, and potential for replication in academic settings.