A Robot Operating System (ROS) based humanoid robot control

This thesis presents adapting techniques required to enhance the capability of a commercially available robot, namely, Robotis Bioloid Premium Humanoid Robot (BPHR). BeagleBone Black (BBB), the decision-making and implementing (intelligence providing) component, with multifunctional capabilities is used in this research. Robot operating System (ROS) and its libraries, as well as Python Script and its libraries have been developed and incorporated into the BBB. This fortified BBB intelligence providing component is then transplanted into the structure of the Robotis Bioloid humanoid robot, after removing the latter’s original decision-making and implementing component (controller). Thus, this study revitalizes the Bioloid humanoid robot by converting it into a humanoid robot with multiple features that can be inherited using ROS. This is a first of its kind approach wherein ROS is used as the development framework in conjunction with the main BBB controller and the software impregnated with Python libraries is used to integrate robotic functions. A full ROS computation is developed and a high level Application Programming Interface (API) usable by software utilizing ROS services is also developed. In this revised two-legged-humanoid robot, USB2Dynamixel connector is used to operate the Dynamixel AX-12A actuators through the Wi-Fi interface of the fortified BBB. An accelerometer sensor supports balancing of the robot, and updates data to the BBB periodically. An Infrared (IR) sensor is used to detect obstacles. This dynamic model is used to actuate the motors mounted on the robot leg thereby resulting in a swing-stance period of the legs for a stable forward movement of the robot. The maximum walking speed of the robot is 0.5 feet/second, beyond this limit the robot becomes unstable. The angle at which the robot leans is governed by the feedback from the accelerometer sensor, which is 20 degrees. If the robot tilts beyond a specific degree, then it would come back to its standstill position and stop further movement. When the robot moves forward, the IR sensors sense obstacles in front of the robot. If an obstacle is detected within 35 cm, then the robot stops moving further. Implementation of ROS on top of the BBB (by replacing CM530 controller with the BBB) and using feedback controls from the accelerometer and IR sensor to control the two-legged robotic movement are the novelties of this work.