Design and implementation of A Bio-Inspired Robot Arm: Machine learning, Robot vision

Abstract

This research highlights a method for remotely controlling a highly complex humanoid robot that relies on the operator’s bio-signals The framework combines machine learning, robot vision, and collision avoidance to ensure safe interaction with the environment. This occurs even with rather noisy control signals, such as the acceleration and electromyography (EMG) data from the operator’s hands. Similar actions, such as reaching for and grasping objects, may be performed with the 5-DOF arm with the use of bio-signals.