Experimental land model of tele-operated underwater backhoe with AR technology

Abstract

Teleoperation technology of hydraulic general-purpose machinery was developed for land applications, particularly at disaster recovery sites. The technology level was significantly improved through practical application at disaster sites such as the Mt. Unzen-fugendake eruption in 1994 and the Usu volcano eruption in 2000. On the other hand, underwater work in harbors largely depends on divers because the scale of work is smaller and manual work is generally more appropriate. Another reason is that the special environments can lead to technological difficulties. Underwater work by divers, however, may be inefficient depending on oceanographic phenomena such as poor visibility and strong tidal currents, and it may also be unsafe because of the difficulty in perceiving and avoiding danger. In addition, long hours of underwater work can lead to a heavy physical load caused by the hydraulic pressure, placing greater demands on the worker than when working on land. In order to solve these issues, we propose the use of Augmented Reality technology for the teleoperation of underwater equipment. Using a laboratory model, we conducted controlled experiments and investigations on the reflection-force sensor mechanism. For the interface section, we fabricated a figure input unit similar in relationship to the front part of a backhoe, and conducted experiments to confirm the operational efficiency of its pointing action. This paper describes the development of an experimental land-based machine