Samara scientists have created a sea-bass-formed underwater robot. The underwater drone “Okyn,” made in the form of a giant sea bass of almost 1 m long, can independently navigate underwater, simulating fish movements, when relocating in the assigned direction, by using its tail fin. The development is equipped with the vision system and has germs of artificial intelligence, which helps to bypass obstacles and detect items assigned to be searched. The robot-fish has already successfully managed the first tests in the pool.
“Our development belongs to the class of autonomous unmanned underwater vehicles of bionic type.

Visually, the device is made in a sea bass form of 85 cm long, this is the real size of a real sea bass, sometimes they can be longer. The fish-shaped hull and the propulsion in the form of a flexible tail fin make it possible to achieve high hydrodynamic parameters: first of all, these are higher indicators of speed, maneuverability and acceleration. A fish-formed device requires less energy to move; moreover, it is much more difficult to be detected in the water, unlike traditional underwater drones of non-bionic type,” told us Evgeny Tatarenko, one of the “Okun” developers, a researcher at the Scientific and Educational Centre for Robotics and Mechatronics of Samara University.

According to him, scientists and engineers have tried to make the “Okun” as similar as possible to a real marine creature; they even painted it for fish. The fish-shaped streamlined body was made of plastic with a 3D printer, with its flexible tail fin to be made of polyurethane. A servo drive and two artificial “ligaments” made of ultrahigh molecular weight polyethylene are responsible for the tail movements.

The underwater drone is powered by several lithium batteries. Inside the waterproof case, there is a basic electronic filling: the motherboard with the microprocessor, the vision system with a video camera, a transceiver, a data storage device and the navigation system consisting of sensors of course, depth and the case inclination angle. Despite the significant number of components, the “Okun” weighs only one and a half kilograms.

“The device is designed to work at a depth of about five meters. Because of its developed algorithms, the “Okun” is able to independently navigate underwater, and differentiate the targets assigned. When developing algorithms, artificial intelligence technologies and neural network self-learning methods are used. Depending on the electronic filling, the “Okun” can perform in future a variety of tasks, for example, such as conducting environmental monitoring, patrolling coastal areas of the sea, assisting in search and rescue operations. Currently, the device has successfully passed its first tests conducted in the pool,” said Evgeny Tatarenko.