This paper discusses the design and implementation of a smart IoT-based security robotic dog for intelligently monitoring surveillance. Traditional surveillance systems suffer from issues such as fixed observation points, blind spots, response delays and poor performance in harsh surroundings. In order to mitigate these drawbacks, a quadrupedal robot was designed, which integrates Arduino Uno microcontroller, ESP32-CAM module, ultrasonic sensors, servo motors and various wireless communication schemes. This robotic dog is designed for automatic movement, obstacle detection, environmental detection, video streaming to a web-browser and control of robot movement. An Arduino board serves as the main controller to maneuver the legs and operate sensors and the ESP32-CAM module is used to perform live video surveillance and provide wireless connectivity. Obstacle detection and the awareness of environment are controlled by ultrasonic sensors which allows the robot to navigate in its surrounding environment. The system was designed, assembled, programmed and tested to verify its capabilities. The experiment was successful and demonstrated that the robotic dog is capable of autonomously navigation, obstacle detection, live video stream and movement in response to sensor inputs. The system developed here proves that the merging of IoT technologies and quadrupedal robotics can enhance security monitoring. This proposed robotic dog offers an inexpensive and adaptable option to intelligently monitor and secure areas. Future work can be extended to incorporate artificial intelligence, cloud service and decision making functions in this robotic dog.