Object recognition is essential in any surveillance system. Being important to identify the entry ofintruders to a home or parking in restricted places, are some actions that are performed during a routine. However, the performance of these systems turns out to be inefficient because they remain static and in many cases their viewing angle turns out to be very limited. Therefore, we believe that a mobile surveillance system would be more efficient in reducing citizen insecurity in some closed housing estates, shopping centers or condominiums. Since being a mobile monitoring system would not have the disadvantagesthat present security systems have. Reason why, in this project is proposed the design and applicationofa patrol robot with autonomousnavigation, which is ableto monitor andtravel a specified route. This pathcan be defined by the user, who when recording a waypoint the robot will move to that location in a relatively straight line, allowing the robot to navigate from one point to another pre-programmed. The navigation of robot will be done by GPS, but not only using this signal. Since, although it is true that this gives us information about the position or speed of a topological form, it does not show us if there are anyobstaclesin the way. For this we seek to merge GPS navigation with a sensor-based one, which collects information and with the help of a microprocessor this information serves for the robot to make decisions, since it is sought to be completely autonomous. Finally, the robot will be able to capture images with the help of a camera, storing them in its internal memory. These images are then sent to the user in real time so that the user can analyze them and detect any anomalyactinthe shortest possibletime.