Multidisciplinary Data Fusion for THz Ray- Tracing

Abstract

The human appearance can be interpreted as a disturbance, convenience or even necessity, for example, in technical applications with human-robot interaction and human-gesture detection. This implies the need for implementing a realistic human digital twin with its positions, gestures and individual movements in ray-tracing simulations. Thus, bringing harmony between simulation and measurement owing to the more realistic assumptions within the simulation environment. In this paper, the human-digital twin combo is primarily re-alized based on two rationales: Firstly, a 3D human model is generated whose individual segment movement is represented by individual trajectories. Secondly, the inertial measurement units (IMUs) are placed on the human body for capturing the human motion estimation. The data fusion between different human postures and ray-tracing is achieved using a self-written script applying the translation and rotation of joint movements to the corresponding segments of its digital twin. Investigations using terahertz (THz) ray-tracing simulations for in-room environment scenario are carried out for different human gaits and velocities at 300 GHz center frequency. Moreover, upon comparing ani-mated postures of this human model with its digital twin which executes the movements according to the measured data from the inertial sensors, the need to record human individual postures is stressed in order to simulate the interaction between humans and technical equipment whereby small changes in human movement might trigger a major impact.