An autonomous navigation system for an unmanned surface vehicle for plastic waste collection

• Autores: Figueroa, Piero, Marsano, Luis, Vinces, Leonardo, Vargas, Dante
• Formato: artículo
• Fecha de Publicación: 2023
• Institución: Universidad Peruana de Ciencias Aplicadas
• Repositorio: UPC-Institucional
• Lenguaje: inglés
• OAI Identifier: oai:repositorioacademico.upc.edu.pe:10757/673066
• Enlace del recurso: http://hdl.handle.net/10757/673066
• Nivel de acceso: acceso embargado
• Materia: Artificial Intelligence; Autonomous Robot; Computer Vision; Navigation System; Plastic Contamination; Unmanned Surface Vehicle

This article proposes the design of an autonomous navigation system intended for Unmanned Surface Vehicles (USVs) in the form of catamarans with differential propulsion, whose primary function is the collection of plastic waste in aquatic environments such as seas and lakes. The fundamental goal of this system is to optimize the operational capabilities of the USVs by increasing their navigational autonomy, consequently reducing the costs associated with their operation and production. In this context, the vehicle has been mathematically modeled, and a navigation system composed of four essential components has been established. The first component involves heading control, which utilizes GPS and IMU to estimate the orientation of the USV. The second component is the obstacle avoidance system, which employs a lidar sensor. The third component is a collection system that relies on a computer vision model with convolutional neural networks to detect different types of plastics and manage the collection process. Finally, an architecture has been designed to facilitate the interconnection of all the aforementioned control systems.