The authors have developed a five-finger robotic hand with artificial fingers consisting of closed linkages and a single planetary gear system. It was designed for usage as a prosthetic hand, mainly focusing on stable envelope grasping. However, pinching an object with some fingers is an important operation of the hand in daily life. Therefore, this paper focuses on the function of human pinching motion, in which the DIP joint passively rotates and the direction of the most proximal link of the finger adaptively changes. The authors consider that this phenomenon may provide stable pinching by controlling the force vector exerted by the fingertip on objects while maintaining contact with them. To achieve the above function, we propose a new finger mechanism equipped with a hyperextension joint, with which the finger gains the ability to control the reaction force vector exerted at the fingertip. First, this paper describes the finger mechanism to explain the necessity to introduce a hyperextension joint from kinematic and kinetic points of view. Second, it verifies them through a simulation study and shows that it allows the reaction force vector to be controlled as a result of introducing the hyperextension joint.

In this paper, we propose a performance planning method using benchmark data in Quality Function Deployment (QFD). The performance planning is a series of study such as verification of the target value, parameter decision, feasibility prediction, etc. before progress to the design phase. In this study, the relationship between parameters and the distribution of data are visualized by displaying network graphs and X-Y graphs using quantitative data measured for multiple samples. Combining these X-Y graphs, efficient planning performance become to be possible. Seat comfort performance is taken up as a case study, then the performance planning method is applied to confirm the effectiveness. As a result, it is considered that this method is effective as a further usage of QFD benchmark data.