Mechatronics Design and Motion Simulation of a Hybrid Wheeled-SCARA Mobile Robot Coupled with Artificial Vision for Solid Waste Collection on Stone Beaches

Plastic on beaches and seas is a problem that affects the entire planet; more than 300 million metric tons of plastic are produced per year, in contrast to the 1.5 million metric tons recorded in 1950. In a world where the annual production of plastic is measured in hundreds of millions of tons, it is essential to critically address the collateral effects of this material on our environment, intending to propose innovative and sustainable solutions to mitigate its negative consequences. This study aims to solve the health problem caused by plastic bottles on beaches. The proposed collector robot aims to be a solution dedicated to beach cleaning by having the capacity to collect plastic bottles from 0.5 to 2 liters. The robot is a ROVER-type mobile, a Selective Compliance Assembly Robot Arm (SCARA) type manipulator, an identification system based on You Only Look Once (YOLO) v8, and a master-slave control of electronic subsystems. The structure of the mobile robot is made of 6063-T6 aluminum and EN-AW 1200, while the manipulator uses T800s carbon fiber. The parts of the mobile structure were subjected to loads until 1000 N, resulting in a maximum deformation of 8.398 μm and safety factors above 2.7, ensuring that the structure support the estimated weight of 60 kg. The control was developed in Matlab with a PID closed-loop system and was validated in tests that confirmed its correct operation. In the tests of the artificial vision models, all 5 bottles that were placed were detected, achieving a minimum match of 0.84 with the nano version of YOLOv8. A virtual environment was recreated to guarantee the proper functioning of the systems independently and intertwined, mobility, identification, and operation tests were carried out, achieving optimal results.