Generation and retrieval of procedural memory using natural intelligence for an articulated robot

This article presents the basics of a working framework to emulate the cognitive function of procedural learning found in natural intelligence to aid in kinematic decision making in a robot. This contrasts with pure iterative recursion or regression techniques. Articulated robots comprise of only rotary joints with a stationary or mobile base reference to the world and are useful in manufacturing, maintenance, and operations to make labor intensive work easier for humans. The motivation for this work is to utilize the potential of cybernetic automation systems to keep up with scaling demands in supply chain production and further build sustainable means to support our growth in the future. In this research a robot is modelled for the process of calculating actuation commands from a starting point in the Euclidean task space to a target. This is known as Inverse Kinematics (IK), named in distinction to Forward Kinematics (FK) which calculates where a point on the robot will be in the task space, for a predefined set of joint positions. We aim to cover the gap in research to connect the development of procedural memory to path planning for a link manipulator robot. We ask the question, if there is any benefit of using procedural memory to reduce the calculation time taken for inverse kinematics to predict a path.