Assistance for Master-Slave System for Objects of Various Shapes by Eye Gaze Tracking and Motion Prediction
Ayaka Matsuzaka, Yang Liu, Chuangyu Guo, Takuya Shirato, Akio Namiki · 2018
In recent years, the demand for robots that can perform various tasks on behalf of people has increased. A master-slave system is a system that can be operated intuitively. However, since complicated operations are difficult for the operator to perform, we need systems that can perform tasks more quickly and with less effort. In our laboratory, we have developed semi-autonomous control (combined operation control and autonomous control) based on prediction of the operator's motions in specific tasks. In particular, reaching and ball catching were realized. However, there are sometimes wrong judgments, and it is difficult to apply this system in the case of multiple tasks. Therefore, it is necessary to predict the operator's intention before his/her motion. In this paper, we propose assistance systems for objects of various shapes by eye gaze tracking and prediction of the operator's motion. Finally, we aim to design and construct a master-slave system that can perform various tasks safely, more quickly, less effort and accurately.