In-plant 3D positioning system using point cloud data for remote decontamination machine

Abstract

Toshiba Corporation, a member of International Research Institute of Nuclear Decommissioning (IRID), has contributed to decontamination works throughout Fukushima Daiichi Nuclear Power Plant from outside ground to inside of the buildings. Speedy decontamination works allow workers to access and stay inside of the contaminated buildings for many hours, and as a result, all decommission works can be accelerated. Some remote decontamination machines to decontaminate the inside of the reactor buildings from a remote safe building have been developed for workers not to be radiated by high-level radiation. Conventionally, operators have remotely controlled the decontamination machine through multiple views of some remote surveillance cameras mounted on it. Because the position data such as GPS data is not available in the buildings, it was hard for operators to detect its absolute position and orientation in the building, and it took much time to recognize targets to be decontaminated. In order to reduce positioning time and make operation works easier, we constructed 3D positioning system to automatically detect the absolute 3D position of the decontamination machine in the reactor buildings from a remote safe building. Moreover, we can also keep records of decontamination works easily by tracking 3D position of the decontamination machine. This paper shows the overview of our approach of 3D positioning and a result of examinations in the mock-up facility that simulates a part of the inside of the reactor building at Fukushima Daiichi NPP.

Keywords

• decontamination
• remote system
• image processing
• 3d positioning
• 3d point cloud