Maqya: Design of an autonomous low cost beach-cleaner robot for endanger beaches

Rivadeneira F.; Martinez S.; Teran A.; Arauco B.; Flores J.; Furukawa R. · Proceedings of the 2023 IEEE 30th International Conference on Electronics, Electrical Engineering and Computing, INTERCON 2023 (2023) · DOI: 10.1109/intercon59652.2023.10326032

Resumen

Pollution is a frequent problem on the most crowded beaches in cities. To solve this problem, the design of a robot integrated with computer vision that identifies places affected by excess waste is proposed and through the mechanical system, the collection of these is carried out for subsequent disposal. Hence, we have developed a low-cost autonomous cleaner robot design that using a YOLO V8 architecture will detect garbage on the sand beach in order to collect them. As a result, the robot was simulated using Robot Operating System (ROS) framework, and the predictive model with the best performance got a 90.95% precision, 78.37% recall, and an F1-Score of 84.19%, using the TACO dataset to train it.

Instituciones: Pontificia Universidad Católica del Perú