Modelagem, controle e realimentação de força para a teleoperação de manipuladores móveis do tipo escavadeira

Abstract

Teleoperation can be defined as the operation of equipment or devices located physically distant from the operator, extending the human capacity to manipulate objects and control equipment remotely, providing the operator with the necessary information to make the interaction as natural and transparent as possible. Such systems can be applied in different areas such as medical surgery, space exploration, handling radioactive materials, inspection of confined areas and mining. In the case of mining, teleoperation can help reduce the physical presence of field workers, increasing safety and comfort for mining equipment operators, as well as enabling the increase in productivity. In this context, the project Teleoperação Avançada de Equipamentos de Mineração, carried out in partnership between Instituto Tecnológico Vale (ITV) and Laboratório de Visão Computacional e Robótica (VeRLab) of the Computer Science Department at Universidade Federal de Minas Gerais (UFMG), aims to develop a teleoperation computational framework to perform remote tasks with an excavator-type vehicle. This master thesis is part of the project and presents a study and investigation of techniques considered as state of the art to compose three framework force feedback configurations: the first configuration without any kind of haptic information from the remote environment; the second with force feedback allowing the operator to sense the excavator’s bucket interactions with objects; the third combining the interaction forces with those that direct the operator to the excavation/deposit site. The different teleoperation strategies are experimentally verified, through tests carried out in a indoor environment using a mobile platform with a manipulator arm, similar to a scaled-down excavator. The experiments consisted of using a haptic device and a joystick to operate the robotic excavator and emulate a material loading/unloading task. Finally, the NASA Task Load Index (NASA-TLX) was used with a group of volunteers to measure the workload when teleoperating the robotic platform using the proposed operating configurations. The results show that the force feedback reduces the equipment’s energy consumption and the physical stress of the equipment during direct contact with a surface, assisting the operator and providing safety to the machine during teleoperation. However, the workload to perform the task, measured through the NASA-TLX, increased for most volunteers, revealing the necessity for labor specialization as the system becomes automated.