Little about human-robot interaction in less technologically advanced communities has been studied. This project seeks to develop a tour guide robot to administer tours to visitors, students, faculty, and staff of Ashesi University College in Ghana. By doing this, it hopes to achieve two things. The first is to bring the discipline of robotics, and more broadly those of computer science and engineering, to the consciousness of students as exciting and worthwhile study and career paths. The second is to study the dynamics of the interaction between humans and robots in the context of an African university community.
The robot tour guide is being developed by students and recent graduates at Ashesi University College, supervised by faculty. The project use a low-cost robotic research platform, the Turtlebot, which comprises of a mobile robot base for locomotion, the Microsoft Kinect for sensing, and a netbook for computation. It runs the Robot Operating System, an open-source Linux based set of software libraries that includes modules providing basic functionality such as locomotion, mapping and navigation.
The architecture of the system is made up of three distinct layers. The topmost layer is an interaction layer that facilitates interaction between the human user and the robot. It allows users to specify the locations in the environment they would like to visit. The intermediary layer is a planning layer that determines the sequence of locations that the robot should visit given the user’s input, and consequently the maps that should be loaded into the robot’s memory at any given time. The lowest layer is the navigation layer that controls the robot’s navigation from a given start location to a goal location.
Currently, the navigation layer of the system has been designed and implemented. With it, the robot can navigate an already mapped environment, visit indicated points of interest and play a recorded sound bite that talks about the location much as a human tour guide would. The next phase of the project involves developing the planning and interaction layers of the system. When complete, the working tour guide will provide rich opportunities for studying human-robot interaction in the context of an African University.