Prediction method of an in-pipe robot's orientation to pass in a curved pipe

Pipe systems are infrastructure of a modern society. Since inner conditions of pipe systems could affect quality of transferring mass, inspecting and maintaining pipe system is an important issue. Many in-pipe robots have been introduced and are under developing to resolve the problem. Pipe systems are composed of different shaped pipes such as straight, curved and branch pipes. Therefore, moving algorithms are introduced. In order to move in a bent pipe, detecting the orientation of an in-pipe robot is an important problem. In this paper, we propose a predicting method of in-pipe robot's orientation, and verify it by experiments.