“Mask-Bot 2i”: An active customisable Robotic Head with Interchangeable Face Brennand Pierce, Takaaki Kuratate, Christian Vogl and Gordon Cheng Institute for Cognitive Systems (http://www.ics.ei.tum.de), Technische Universit¨ at M¨ unchen Abstract—This paper describes the development of “Mask- Bot 2i” (codenamed: Kabuto), a robotic head designed for research into Human-Robot-Interactions. The uniqueness of our new robotic head is that the appearance of its face can be alter on-the-fly. Different faces can be projected onto the active head system. The head is fully active with a 3-DOF neck, with support for biannual hearing as well as video camera for seeing. An interchangeable face is the main feature of this new Mask-Bot, the head can be equipped with an average face mask as well as a highly customised individualised face that can be easily exchanged. Additionally, the actuation of the head has been designed to match the natural head movements of an average human. Thus, enabling the head and face to be articulated synchronously to the speech production while the natural head movement matches that of the animated face. The design and realisation of this new system is presented in details in this paper. I. I NTRODUCTION During the last decades numerous humanoid robotic sys- tems have come to be. One element that is consistent among them all is the fact they are all equipped with a head. As from the practical point of view, it is the most logical place to mount the visual and the auditory systems. But one might say that it is also an important part of human-robot interaction (HRI), especially for face-to-face communication, as the head provides a natural means of human interaction with a robot. This leads to the question, “what elements are important in the design of a robotic head for HRI?”. For instance, we have studied how the quality of the projected image effects people’s identification of the avatar’s gender with the original “Mask-Bot” [1]. For interaction, the head should be able to communicate using auditory (verbal and non-verbal) as well as visual communication. But what about its appearance? To explore this question, we developed a robotic head that can be used to carry out research into how the appearance of an animated face displayed onto a robotic head effects HRI. When a robotic head is designed it normally has one primary goal. For example, it could be designed with the aim of being a stable platform for activated cameras. Good examples of this are the head of the “iCub” humanoid robot [2]; or Karlsruhe humanoid head [3]. Another aim is to look as life-like as possible, where the texture of the skin, the mimicking of the muscles are most important. Hanson robotics developed a very realistic head used in the “Albert” version of HUBO humanoid robot [4], which utilises a high number of servo motors that were designed to mimic human muscles and deform its rubbery skin. This means the head is able to display emotions as well as to articulate the mouth Fig. 1. The new “Mask-Bot 2i” (codenamed: Kabuto). when it speaks. Simpler heads compared to the Hanson’s head are also used for displaying emotions. An early example of this type of head is MIT’s KISMET robot [5], which was designed to display emotional expressions and had a simple mouth. These latter two examples rely heavily on complex mechanical structures, which need a high number of motors to be controlled in order to modify their facial expressions. To overcome these problems, there is an emerging type of humanoid heads with the concept of displaying an avatar instead of relaying on a complicated mechanical mechanism. Examples of this would be the “LightHead” by Delaunay et al.[6], [7] and the “Curved Screen Face” from Hashimoto and colleagues [8], as well as our own robotic head “Mask-Bot” [9], [10]. This type of robotic heads have the advantage that the face can be changed and animated fairly easily. Also the articulation of the face does not rely on complex mechanical components. This means that the mouth can be animated and synchronised with the vocal system and it can also display emotions. But current research on the projected heads have mainly focused on the animated side and have neglected the complete modalities that a robotic head needs, for ex- ample the integration of stereo microphones for “hearing”, cameras to “see” the world and the full articulation of head 2012 12th IEEE-RAS International Conference on Humanoid Robots Nov.29-Dec.1, 2012. Business Innovation Center Osaka, Japan 978-1-4673-1369-8/12/$31.00 ©2012 IEEE 520