Development of the robot balanced on a single ball -- Motivation, Approach, Implementation --
Talk 4:30 pm
Another version of the ballbot that I developed with my students is the main topic of this talk. The project started from my students' desire to build such a robot. The projects in my laboratory, Robot Development Engineering Lab., usually started in this way. Most of the projects were not research work but rather development training for the undergraduate students, who proposed many interesting ideas that I would not have conceived of. Due to their lack of formal training, they are not influenced by popular robot design or similar previous work. This can lead to their proposing strange new ideas. Our ball-balancing robot, BallIP, employed a three-dimensional ball drive mechanism using a kind of wheels for omnidirectional robots and a simple inverted pendulum control. By using the acceleration inputs to the system, we enhanced the robustness with respect to an increase of the system's inertia, which is useful for when the robot is carrying something. Some of the latest results obtained after ICRA'09, including a new driving mechanism, will be briefly presented. This talk will not be theoretical nor analytical but technical. As the background of the work, other projects on biped walking robots with roller skates and several amusing robots will be also presented.