Drone Reconfigurable Architecture (DRA): a Multipurpose Modular Architecture for Unmanned Aerial Vehicles (UAVs)

• Autor(a) principal: da Silva Ferreira, Murillo Augusto [UNESP]
• Data de Publicação: 2020
• Outros Autores: Begazo, Maria Fernanda Tejada [UNESP], Lopes, Guilherme Cano, de Oliveira, Alexandre Felipe [UNESP], Colombini, Esther Luna, da Silva Simões, Alexandre [UNESP]
• Tipo de documento: Artigo
• Idioma: eng
• Título da fonte: Repositório Institucional da UNESP
• Texto Completo: http://dx.doi.org/10.1007/s10846-019-01129-4; http://hdl.handle.net/11449/200558
• topic: Aerial robots; Drones; Modular robots; Reconfigurable robots; UAV

This work proposes the Drone Reconfigurable Architecture (DRA), which is a modular architecture for UAVs with electrical, mechanical, and computational specifications. The theoretical aspects of the architecture are introduced through a case study with practical implementations aiming to design a multi-rotor UAV, which also includes the manufacturing steps of a functional prototype. Our proposal can be used in a scenario where the capacity of physical reconfiguration of a UAV would confer an enormous advantage to these aircraft in terms of applicability. This happens in the case where each task typically requires a robot with a particular physical architecture (number and position of propellers, autonomy, thrust, sensors, and communication). Results of a set of tests with an aircraft assembly are presented to verify the versatility of the proposed architecture, demonstrating the better performance of these aircraft when compared with conventional UAVs. The proposed methodology allows applications in a variety of scenarios like cargo transportation, support, agriculture, publicity, pest control, surveillance, inspection, and entertainment, between others. In these scenarios, although a software with some generic components could easily control drones to perform all of them, it is unthinkable to consider that a single drone with a particular physical structure would be able to be adapted to all of the tasks necessary (as path following, localization, and mapping).