In the development of medical skill training systems, the efficiency of the system and the provision of quantitative feedback information to the trainee are very important. Furthermore, usage of the simulated operation platform should be as realistic as possible. In order to satisfy these requirements, we developed a robot to be used for airway management training: Waseda KyotoKagaku Airway No.1 Refined RII (WKA-1RII) (Fig. 1). In addition to realistically shaped hardware and various sensory equipments inside the robot, the new training system also uses a binocular vision system, an inertial measurement unit and a 3D simulation software component to track the movements of the trainee. For fully estimating the skills of the trainee and providing richer feedback, it is not enough to only use information about the tool's movements from the inside of the robot. Therefore, we propose a system to fuse the sensory information from inside the robot with data from 3d vision and from the inertial measurement unit. This accurate information about the movements of the trainee is used to model the progress of the training with a 3d computer graphics simulator. The trainee can use this visualization during or after the training procedure to verify his training status. Using this system he has the possibility to easily compare his performance with a guideline performance provided by an experienced surgeon. In this paper we show a first conceptual application of this approach. The experimental results lead to the consideration that the approach is worth following in further research.