Conceptual Development of an Autonomous Underwater Robot Design for Monitoring and Harvesting Invasive Weeds

Abstract

The design of a biomimicry autonomous underwater robot for monitoring and harvesting invasive weeds in lakes is presented in this paper. The systematic design of the robot focuses on integrating 5G-AI-IoT as effective technological tools to autonomously monitor and harvest invasive weeds in order to replace traditional weed control approaches. The robustness and versatility of the robotic platform to structural topology and autonomous navigation that uses convolutional neural network methods and unsupervised learning techniques will be demonstrated. The robotic concept design will investigate real time sensing, mapping and visualization of the invasive weeds. The system based on real-time mapping information obtained from the swarm of drones will also manage the control of the underwater robots equipped with smart networked sensors using State of the Art IoT technologies. The mechanical dislodging machine will be guided to the mapped areas and accurately controlled and guided through smart sensors via the 5G Ultra-reliable Low-Latency Communication Control (URLLC) and tactile control system to dislodge the invasive weed with no impact on other organisms and the biodiversity of the lake.