Comparison of remote and local virtual fixture and method to improve performance of them

Abstract

Recently, to improve the performance of a teleoperation, virtual fixture that provides virtual force as haptic feedback has been widely researched. Although, there are two types of virtual fixture depending on their generation method, Many researcher has used without classification. Because virtual fixtures improve performance without classification. However, since virtual fixture has different force feedback characteristics depending on the generation method, there is optimal virtual fixture type that shows better result according to purpose. Therefore, in this paper we find which virtual fixture is better to use for each purpose by comparing remote and local virtual fixture through experimental task. In addition, we proposed method to improve performance of them according to ideal virtual fixture condition. In many previous research, virtual fixture are predefined and do not consider time delay in communication channel. However, actual teleoperation environment is not structured, which cannot be defined virtual fixture in advance, and there is an time delay. Therefore, we presents a algorithm to reduce the position error of the virtual model that can occur when creating a virtual guide using a camera and the force output error that can occur when there is a time delay. More specifically, we update the virtual model in real time with the force/torque data that occur when the robot contact with the environment and reduce the time delay effect through frame transformations within the model. The proposed methods have been evaluated through experiments and confirmed that it can be successfully improved.