Abstract: Driving vehicles is one of the amusing things that individually do, a side interest for another but in the meantime is a hazardous instrument that could lead to death if used in the wrong way. Recently, the number of vehicle’s theft had risen essentially which led to great losses at both individuals and establishment (insurance agencies) levels. Thus, a great interest emerged as ways to protect vehicles from theft. This study presents anti-robbery and protection framework that consists of two phases. In the first phase, face recognition techniques are employed to identify the vehicle’s owner to secure the vehicle from burglary. Once recognized, the bluetooth on the phone activated and signally connected to the arduino which is fitted inside the vehicle. Then arduino connecting (shield, wires, servo motor) of the vehicle door lock and engine; opens the lock and start the engine. Otherwise, if the user has not identified the door remains locked and the alert system is activated. In the second phase, during driving the eye locomotion is observed using a mobile to reduce or limit collision. In the case of proven snoozing, the reflexes of the driver are translated by the designed system to warn the driver by a beep to save his life and also signal an alarm that stops the vehicle sequentially using its brake system. The process is repeated every 6 sec to perceive if there is any distraction or somnolence that might occur during driving or not. Three types of the database have been used to test the proposed framework namely, face 94, ORL and Live database. The performance evaluations metric based on the False Rejection Rate (FRR), False Acceptance Rate (FAR) and elapsed time have been used to assess the effectiveness of different facial recognition techniques. Software results using MATLAB on a test set of photos have proved that the Principle Component Analysis (PCA) technique has a superior performance. The executed framework uses the android operating system in a smartphone to assist in detecting drivers under fatigue and alert driver under sleepy conditions. Comparison different techniques, the Block Matching Algorithm (BMA) demonstrated a superior performance for tracking driver’s eyes for limiting a collision in real-time according to elapsed time parameter. Hardware implementation is executed using mitsubishi lancer vehicle and smartphone (Samsung Galaxy S4) has been used for testing the proposed framework.