A Study on Window-based Path Driving through Identifying the Target Walking in an Ultrasonic Sensor-based Follow-up Collaborative Robot
Seung-Ho Yum, Na-Yeon Park, Jinwoo Shin, Su-Hong Eom, Eung-Hyuk Lee · 2022
This study proposes a method for window-based path driving through identifying the user's walking and works in an ultrasonic sensor-based follow-up collaborative robot. Target location estimation methods of an ultrasonic sensor-based follow-up collaborative robot are divided into distance-based and coordinate-based following methods. It is difficult to smoothly use the distance-based following method in an unspecified indoor space due to the time delay in its following process since it follows a target with a safe distance. A preliminary study that converts the distance into location information in the target following process in order to compensate for it causes fails in the following due to frequent path changes when errors occur in reference distance values because of irregularities such as specular reflection in an ultrasonic sensor. Therefore, this study proposes a window-based following method according to the identification of user's walking and works in a robot through converting the current location of the robot into a calibrated coordinate system. The driving of the proposed system is based on encoders attached to the robot and the cumulative error is minimized by initializing the coordinate in the user's identification of works to prevent errors caused by slip due to long-term operation.