Simulator for learning symbolically about the behavior of motion in bipedal robots

The main objective in developing the simulator presented in this paper is to assist undergraduate students to learn about the analysis of motion behavior in bipedal robots by exploring an approach based on fuzzy-symbolical variables. This approach facilitates learning about and understanding of motion behavior using symbolic reasoning instead of numerical reasoning, such as that based on the analysis of moments and forces, which are difficult to grasp. An important educational aim of this simulator is to develop cognitive skills such as analysis, the construction of structures and generalization. In addition, the simulator helps students to develop an aptitude for self-directed learning as the students themselves build the rules that help them to understand the behavior of motions. The simulator has been built to interact with two real bipedal robots that have 12 degrees-of-freedom (DOF). Experimental results showed that the students were able to understand and analyze motion behavior and then predict the risk that the robot would fall as they interacted with the simulator by reasoning using symbolical variables and their relations. In addition, students developed their cognitive skills throughout the learning process.