Virtual Booth

a.     The goal of the underwater drone is to create an underwater drone that possesses the ability to detect objects in real-time using YOLOv7-tiny model, which has been tuned for rapid processing. The drone is powered by NVIDIA jetson Nano platform, which is a key component enabling efficient processing of deep learning algorithms for object recognition and navigation tasks. To ensure a portable design that integrates Arduino Uno for control and Jetson nano for the real-time data analysis is implemented. The drone is designed to be waterproof, featuring a junction box with a IP67 rating and sealants. This design allows the drone to operate seamlessly in underwater environment, making it exceptionally suitable for a wide range of applications including marine research, environmental monitoring, underwater inspection, and exploration ventures. The achieved accuracy in object detection spans from 60% to 85%. The drone manages to achieve detection speed ranging from 7 to 15 frames per second, which satisfactorily meets the real-time demands posed by practical underwater applications. Moreover,  the drone has capacity to acquire water samples intended for thorough analysis. The drone is capable to transmit the gathered data to a surface vessel or a station located on the shore, enabling continuous monitoring and in-depth analysis. This capability facilitates swift response to potential threats and plays a crucial role in safeguarding the ecosystem and marine organisms.