AI-Assisted Intelligent Humanoid Robot

One of the most difficult tasks in robotics is to replace one human-operator by an intelligent, autonomous and emotional humanoid robot for conducting various types of real-world tasks. To perform a pre-specified task, the optimal design of a proposed humanoid robot is being frozen after carrying out its kinematics and dynamic analysis. Based on this optimized design, the robot is fabricated, and its performances are tested in the Laboratory. If it is found to be satisfactory, it is then used in the real field. Let us assume that the developed humanoid robot is facing a few undesirable disturbances due to some changes in the working environment. For example, one humanoid robot working as a soldier in the real field suddenly faces some un-expected rough terrains to be negotiated by maintaining its dynamic balance and consuming minimum energy. This paper deals with the issues related to negotiating these sudden changes in the environment and varying loads with the help of its AI-assisted adaptive vision system, multi-sensors data fusion system, motion and gait planning schemes, controller, and so on. After summarizing this study, some scopes for the future study have also been suggested. To make an intelligent humanoid robot capable of negotiating the varying situations, it should be equipped with AI-assisted vision system, adaptive motion and gait planners, adaptive and robust controller, and others.