This article aims to test a facial recognition-based front door security system, which can also convey relevant information to the owner’s mobile phone via an SMS gateway. This system is necessary to prevent unwanted criminal activity by the owner. The method used is Hercascade as face recognition for security. Use a set of webcam settings to compare human face objects in the background with face data already stored in the database. Capture images using a Raspberry PI connected to a USB webcam for the sensor, move the front door using a servomotor as a drive, and own the system in the form of a short message from basic communication process data Notify to. There is also an ultrasonic sensor as an activation system to detect human objects when they approach the door and invade. Based on the results of 90 tests on a system with varying distances of objects on the camera (30 cm, 40 cm, 50 cm), the average pass rate of the tests is 91.11%. We can conclude that face recognition by the Hercassette method can be applied as an entrance security system.