Home /Research /Human-robot interaction methodology
HRI

Human-robot interaction methodology

A.S. Sekmen, Mitch Wilkes

Year
2000
Citations
11

Abstract

This thesis investigates different phases of human-robot interactions. The starting point is to classify human-robot interactions into three modes called passive, semi-active, and full-active human-robot interactions. The differences between these modes are presented, possible contributions are discussed, and some of the modes are tested on different service robots. In addition, the importance of relative locations and prior knowledge is investigated. In passive interaction, a human and a robot directly interact without a medium such as a joystick or a user interface. A human is passive in the sense that he/she does not need to direct the robot by sending specific messages through a communication means such as a computer or a joystick. The robot has to extract specific features from the physical situation of the passive user. Two new methods, which are human tracking by sound localization and infrared motion detection respectively, are implemented on a human service robot called ISAC. The human can be considered as a passive user since he/she behaves naturally. In the active interactions, the human is active in the sense that he/she needs to direct the robot to some extent. Remote shopping trips can be considered as a potential application. Two control mechanisms, manual control and autonomous navigation, are developed and implemented on a mobile robot called TourMate. These mechanisms allow a user to control TourMate over the Internet. This interaction can be considered as semi-active interaction in the sense that the robot does not learn from the user. In full-active interactions, the user does not only learn from the robot but also can teach it about the environment. For passive interactions, the robot and human are in immediate locations, however, they may be in immediate, intermediate, or remote locations for active interactions. This dissertation investigates the importance of being in the intermediate or remote locations for active interactions. The theory and procedure used to develop a human-robot interaction methodology are included in this dissertation. The experimental setup and procedure used to conduct the experiments as well as the results of the experiments and its analysis are all reported in this dissertation.

Keywords

RobotJoystickHuman–computer interactionSocial robotHuman–robot interactionComputer scienceMobile robotRobot controlPersonal robotUbiquitous robot

Related papers

Browse all HRI papers