Abstract

This paper introduces an innovative approach to underwater exploration by integrating Artificial Intelligence (AI) into Autonomous Underwater Vehicles (AUVs). This collaboration between AI and biomimicry marks a new era for AUVs, enabling them to emulate marine creatures’ graceful and efficient movements. By infusing AI capabilities into AUVs, AUVs are empowered to learn and adapt, making autonomous real-time decisions without human intervention. This dynamic integration equips AUVs to effectively navigate complex underwater terrains, evade obstacles, and seamlessly interact with marine life. Inspired by the remarkable propulsion mechanisms found in marine organisms, this work proposes a pioneering propulsion system tailored for AUVs. Taking cues from the locomotion of creatures like cuttlefish, the biomechanics is translated into a robotic propulsion system. The result is a fluid and energy-efficient propulsion method that mitigates the harmful effects of cavitation, thereby reducing noise pollution and minimizing disruption to marine ecosystems. This research evaluates the performance of on-device AI models for analyzing the sensing environment around the AUV and taking real-time images. This automated sensing and navigation method can help the AUVs independently navigate to the desired location along the water table. The propulsion is achieved by building a crankshaft mechanism and a unified mechanical design to convert rotational motion from a motor into a sinusoidal wave motion to replicate the cuttlefish locomotion pattern. The proposed underwater vehicle, Aquabot, is designed using Fusion 360 simulation and ANSYS software. The results demonstrate the accuracy and efficiency of the autonomous underwater vehicle based on the environmental conditions, thus reducing energy consumption and enhancing aquatic vehicle efficiency.

Description

2024 The Authors. This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License. For more information, see https://creativecommons.org/licenses/by-nc-nd/4.0/

Publisher

IEEE

Date of publication

4-2024

Language

english

Persistent identifier

http://hdl.handle.net/10950/4722

Document Type

Article

Share

COinS