Hyper realistic digital twin environments are limited to demanding computational requirements of running accurate simulations. More so when these environments are meant to be rendered in real-time for training and XR. In order to address this problem, Buendea in collaboration with the University of Central Florida’s Institute for Simulation and Training, are proposing the development of a system that enhances real-time XR capabilities for astronaut training through the off-loading and processing of demanding simulation tasks to cloud-based virtual machines.
By enabling parallel processes that run side-by-side to real-time simulations, we intend to layer microservice extensions that produce more accurate simulation models required in XR astronaut training. This effort will also demonstrate that this approach can be expanded to other parallel simulations that enhance accuracy of a digital twin without affecting real-time simulation rendering performance of an XR environment.
This effort will be focused on the development of a persistent simulation layer that provides ongoing critical data for evaluation of EVA training scenarios on the Moon and Martian surface. Multi-GPU virtual machine instances will run in parallel to local XR environments and provide real-time simulations of critical mission processes, human performance, and environmental conditions. Data from these sessions will be stored and accessible for review to help improve operational XR training capabilities and simulation accuracy.