Application of Deep Learning Model for Analysis of Forward Kinematics of a 6-Axis Robotic Hand for a Humanoid

Abstract

Rigid body kinematics can be generically categorized as forward and inverse kinematics. Forward kinematics calculates the precise location and orientation of the tool at the end of a robotic arm based on the specific angles of its joints. In contrast, inverse kinematics calculates the necessary joint movements to reach a specific location of the end effector (tool). These kinematics formulations are evaluated using homogeneous transformations. Computation of the solutions for the homogenous formulations can be computationally expensive for forward kinematics. Further, the computational requirements are significantly larger for inverse kinematics. The precision, resolution and accuracy of the robot kinematics depends on complex mathematical formulations. In today's dynamic world, the application complex mathematics is sometimes considered as exhausted. The increasing deployment of artificial intelligence tools and machine learning technologies is heading towards a paradigm shift towards advanced frameworks with less complexities. This research article presents an innovative artificial intelligence (AI) method that deploys an advanced deep learning techniques to precisely forecast the forward kinematics for a specifically developed 6-axis robotic hand for humanoids.