#TITLE_ALTERNATIVE#

• Main Author: FADLI (NIM: 23216118), HILMI
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/27875
• Institution: Institut Teknologi Bandung

Robot with walking locomotion type, especially humanoid (bipedal), is still very much researched and developed for its flexibility in reaching a variety of terrain and also the similarity with humans. This increases the passion of researchers to develop this robot as assistance robot. However, the number of DOF has become a major problem in the development of this robot. Controlling the robot's motion to get the desired body posture is quite difficult if done manually. In this paper, the authors develop a system to control humanoid robot motion through human gesture imitation. This system allows us to easily instruct the robot to imitate what we’re doing. This system uses Kinect as the sensing device that will recognize human body and capture information about the human gesture. Kinect is a device developed by Microsoft that is used to recognize human body and generate 3D skeleton model by combining information from depth sensor and RGB camera. The obtained 3D model will be processed and translated to a set of motion instructions consisting of the joint rotation angle data. The author use inverse kinematics approach to translate this 3D model data by inversing kinematics equation based on Denavit Hatenberg method to get the related value of rotation angle. This method requires initial position and final position of the joint to calculate how much the joint rotates. Then, NAO humanoid robot will get instruction to move based on obtained angle data to perform the imitated body posture. NAO is a humanoid robot developed by Aldebaran Robotics. NAO equipped with actuators in the form of brush DC coreless with 25 DOF in all body parts. This robot is also equipped with various sensors including cameras, microphones, sonar sensors, tactile sensors, gyrometer and accelerometer. Both Kinect and NAO provide application programming interface (API) which available in several programming language especially .NET Frameworks. All modules in this system implemented using C# programming language. As the result, NAO can imitate human arms gesture, although there is some delay between posture changes. There is still a problem for legs gesture imitation because of balance issue. This system needs balancing management system to provide balance posture for legs gesture imitation. However, the system is expected to contribute to the motion control of humanoid robots, especially NAO.