Design and Development of an Autonomous Drone equipped with AI

We designed and constructed an autonomous unmanned aerial vehicle (UAV) as our final year project. The airframe was built using 3mm Depron sheets, 5mm balsa wood, and carbon fiber strips, ensuring a lightweight yet durable structure. The UAV was powered by a 3-cell, 2200mAh Li-Po battery, which provided an endurance of approximately 15 minutes. For ground support, the landing gear was fabricated with rubber tires and an aluminum undercarriage, capable of absorbing harsh landing shocks while reliably supporting the UAV’s weight. For autonomous flight, we integrated a Pixhawk PX4 flight controller paired with a Ublox M8N GPS module. The GPS continuously provided positional data to the flight controller, enabling the UAV to navigate and reach pre-programmed waypoints. All waypoint planning and calibration tasks were carried out using Mission Planner software. The project also incorporated an artificial intelligence component through the use of ultrasonic sensors mounted at the front of the UAV. These sensors detect obstacles within a range of six feet. Upon detection, signals are processed via an Arduino microcontroller, which commands the servos connected to the flaps. The flaps extend to increase lift, causing the UAV to ascend and avoid the obstacle ahead. The system responds instantly, ensuring there is no in-flight delay. Although the parameters can be adjusted, the maximum stable detection range of the ultrasonic system without erratic servo movement was determined to be 12 feet. Additionally, the UAV was equipped with an FPV (First Person View) camera mounted behind the ultrasonic sensors. To prevent overheating issues commonly faced by video transmitters, the video transmitter was mounted externally on the airframe, ensuring reliable live video transmission during flight.