Title: Development of a walking-trajectory measurement system
Authors: Nina Tajima; Koichiro Kato; Eriko Okada; Nobuto Matsuhira; Kanako Amano; Yuka Kato
Addresses: Graduate School of Mechanical Engineering, Shibaura Institute of Technology, 3-7-5 Toyosu, Koto-ku, Tokyo, Japan ' Graduate School of Mechanical Engineering, Shibaura Institute of Technology, 3-7-5 Toyosu, Koto-ku, Tokyo, Japan ' Department of Engineering Science and Mechanics, Shibaura Institute of Technology, 3-7-5 Toyosu, Koto-ku, Tokyo, Japan ' Department of Engineering Science and Mechanics, Shibaura Institute of Technology, 3-7-5 Toyosu, Koto-ku, Tokyo, Japan ' Division of Mathematical Sciences, Tokyo Woman's Christian University, 2-6-1 Zempukuji, Suginami-ku, Tokyo, Japan ' Division of Mathematical Sciences, Tokyo Woman's Christian University, 2-6-1 Zempukuji, Suginami-ku, Tokyo, Japan
Abstract: In Japan, where the labour force has decreased in recent years, robots have got a lot of attention to supplement the workforce. In the reception and response control system using the robot 'ApriPoco™', measuring the walking trajectory using a laser rangefinder has been studied, but it cannot measure accurately because the trajectory overlaps. In this paper, we propose a system that predicts trajectories in real-time using Gaussian process regression. We experimented by checking the operation of the non-real-time trajectory prediction system. The proposed system improved the probability of obtaining correct data from about 7.4% compared with the previous system. Based on this result, we conducted a demonstration experiment with a real-time prediction system. As a result, the number of coordinate data points was reduced, and the accuracy decreased. We aim to improve the accuracy of the real-time prediction system in the future.
Keywords: laser range finder; Gaussian process regression; trajectory measurement; the number of people; human-robot interaction.
International Journal of Mechatronics and Automation, 2022 Vol.9 No.3, pp.113 - 122
Received: 15 Sep 2021
Accepted: 11 Nov 2021
Published online: 04 Jul 2022 *