The emerging field of medical robotics is aiming in introducing intelligent tools. More recently, thanks to the innovations on robot technology (RT), advanced medical training systems have been introduced to improve the skills of trainees. The principal challenges of developing efficient medical training systems are simulating real-world conditions and assuring their effectiveness. Up to now, different kinds of medical training devices have been developed which are designed to reproduce with high fidelity the human anatomy. Due to their design concept, the evaluation of progress of the trainees is based on subjective assessments limiting the understanding of their effectiveness. In this paper, we are presenting our research towards developing a patient robot designed to simulate the real-world task conditions and providing objective assessments of the training achievements. Due to its complexity; in this paper, we are presenting as a first approach the development of the Waseda-Kyotokagaku Airway No. 1R (WKA-1R) which includes a human patient model with embedded sensors in order to provide objective assessments of the training progress. In particular, we have proposed an evaluation function to quantitatively evaluate the task performance by determining the weighting of coefficients. In order to determine the weighting of coefficients, we applied discriminant analysis. In order to determine the effectiveness of the proposed evaluation function to detect differences among different levels of expertise, an experimental setup was carried out. From the experimental results, we could find a significant difference between both groups (P<0.05).