A Study on Different User Interfaces for Teaching Virtual Borders to Mobile Robots

Human-aware robot navigation is an essential aspect to increase the acceptance of mobile service robots in human-centered environments, e.g. home environments. Robots need to navigate in a human-acceptable way according to the users’ conventions, presence and needs. In order to address the users’ needs, we employ virtual borders, which are non-physical borders and respected by the robots while working, to effectively restrict the workspace of a mobile robot and change its navigational behavior. To this end, we consider different user interfaces, i.e. visual markers, a laser pointer, a graphical user interface and a RGB-D Google Tango tablet with augmented reality application, to allow non-expert users the flexible and interactive definition of virtual borders. These user interfaces were evaluated with respect to their correctness, flexibility, accuracy, teaching effort and user experience. Experimental results show that the RGB-D Google Tango tablet as user interface yields the best overall results compared to the other user interfaces. Apart from a low teaching effort and high flexibility and accuracy, it features the highest user ratings acquired from a comprehensive user study with 25 participants for intuitiveness, comfort, learnability and its feedback system.