Preprocessing of an Object Transfer Point (OTP) coupled with pre-trained Dynamic Motion Primitives (DMPs), allow for a computationally efficient method to dynamically define interaction space that enables a robot to predict object goal locations for legible Human-Robot Interaction. Report