We present an advanced robot education curriculum, “IT Triadic—RT (robot technology) course” for graduate students of NAIST. The goal of RT course is to educate students so that they acquire an ability to solve challenging and novel problems, such as “Create a robot helpful in the real world.” For this purpose, we design a curriculum centered on project based learning (PBL) where the students propose their own project, apply to get fund, proceed the project, and evaluate the results. Since such a task is very hard to many students, we provide several levels of PBL so that the students can gradually learn how to propose a project; for example, we have brainstorming sessions, and we teach how to write a proposal. In this paper, we describe the curriculum of RT course, its features including the PBL classes and the on-line text system, and the lessons obtained through actually conducting the courses.
This paper presents a shoulder joint for a human-like robotic arm. This joint mechanism is composed of two ball joints laying back to back and the range of motion is larger than that of a usual ball joint. Since the joint is driven by the mutually inter-connected air cylinders, the control signal can be operated equally with a normal ball joint shoulder. The prototype of the joint with 6 air cylinders and its kinematic model were developed to confirm its range of motion. We also achieved an overhand throwing motion with a robotic arm using the proposed shoulder mechanism thanks to the large moving range.
This paper describes a texture sensing system for gel food. Considering the elastic deformation of human tongue, we propose a method to evaluate the food texture by using a simple artificial mastication model with an imitation tongue. After explain a basic principle of variable elasticity of the developed imitation tongue, we show that it can reproduce the elasticity of human tongue from the relaxed to contracted conditions. In addition, we show how the pressure distribution during the artificial mastication of gel differs with respect to the elasticity of imitation tongue. Based on the numerical model between the pressure distribution data and the value of sensory evaluation, we evaluate the food texture. Through the experiment of food texture estimation, we reveal that there exists the appropriate elasticity of imitation tongue, for texture terms and characteristics of gel food.
For the quality-check process of manufacturing, skilled workers often use their sense of touch for inspections. However, this method is non-quantitative and depends on the experience of the workers. Therefore, a measurement device that can acquire quantitative surface profile data more easily than existing equipment is required. In our previous study, we prototyped a cantilever structured thin plate tactile sensor, with an attached piezoelectric polymer film as a strain detector, and showed that this sensor can evaluate uneven surface profiles. However, these measurements were performed by moving the tactile sensor with an automatic linear stage. To make the measurement tasks at an actual work site easier, an operator must be able to choose any target surface and evaluate that surface without troublesome settings and jigs. Therefore, in this study, we prototyped a portable, manually operated tactile sensor system comprising two thin plate tactile sensors. We showed that this prototype sensor system can estimate the amount of sensor movement during measurements by calculating the time lag between two sensor signals. Using the estimated amount of movement, this sensor system can evaluate uneven surface profiles easily and quickly.