About Research

Search for PARI Report/Technical Note

Vision Supplement for Grasping in Unclear Underwater Using AR

Publication year Port and Airport Research Institute Report 041-02-06 2002.06
Author(s) Makoto UTSUMI,Taketsugu HIRABAYASHI,Muneo YOSHIE
Department
/Divison
Construction and Control System Department Control System Division
Executive Summary

 In construction of harbor, most works depend on divers. These works are very hard and dangerous.Considering the efficiency and safety, it is necessary to introduce underwater construction machine.In these present circumstances, installing various type attachments to underwater backhoe increases a wide use of underwater works and develops a mechanization of underwater construction.Especially, teleoperation underwater backhoe with grasping system is useful for setting and removing of brocks or rubbles, dismantlement of broken facilities.
 On the ground, it is able to use a video or eyes vision for teleoperation. However, in underwater, muddiness caused by raising earth and sand disturbs vision. It is difficult to recognize the shape and the center of gravity (COG) of a structure. The efficiency of a grasping operation in water deteriorates.
 Therefore, it is considered that the Augmented Reality (AR) method is applied to the grasping system. In AR method of this research, plural information that is extended or processed compensate for lack of visual information. For this compensation, reaction force and tactile sense is selected and force feedback bilateral control master-slave system is constructed. Reaction-forces and tactile sense give the operator a feeling of grasping. Furthermore, utilize these forces for visualization of haptic image. When ringers of the hand touch a structure, a polygon at the touched point is drawn on the display. This polygon is called Haptic image object. Using this Haptic image object, the operator can understand the shape and COG of the structure easily.
 It became clear that the visualization of haptic image is effective for the grasping operation in poor vision by examinee experiments in a computer simulation. And we developed a master-slave hand models for a next step experiment. In this paper, we present AR method and a validity of this teleoperation system in underwater.

PDF File /en/pdf/en/vol041-no02-06.pdf