UCF presents human digital twin technology at I/ITSEC 2024

By Christa Yates, Staff Writer

 

The University of Central Florida (UCF) presented its human digital twin (HDT) technology during, “A Human Digital Twin Architecture for Knowledge-Based Interactions and Context-Aware Conversations,” a paper presentation held at the 2024 Interservice/Industry Training, Simulation and Education Conference (I/ITSEC).

 

Azhar Ali Mohammad, a second-year graduate student in computer engineering at UCF, presented the HDT architecture as a part of the “Emerging Concepts and Innovative Technologies (ECIT) 9: Foundation Models and Human-AI Interaction” discussion.

 

According to the paper, recent developments in artificial intelligence (AI) and machine learning have provided opportunities for human-autonomy teaming (HAT) to accomplish tasks, missions or continuous, coordinated activities. The goal is to address the challenge of providing awareness and control to humans as they oversee autonomous assets and their actions, as well as having trusted interactions with autonomous assets and supporting HAT as it completes tasks.

 

Mohammad explained that the real-time HDT presented at I/ITSEC integrates large language models that are focused on knowledge reporting, answering and making recommendations into a visual interface that provides a life-like physical embodiment of the autonomous system.

 

“This [technology] faces significant challenges, particularly in establishing trust, maintaining control, and ensuring situational contextual awareness between humans and their AI team members in these systems,” Mohammed said.

 

UCF representatives demonstrated the HDT technology at their booth on the I/ITSEC show floor throughout the conference, wherein the HDT assistant guided participants through the assembly of a model weapon. The AI HDT used cameras to “see” what participants were doing with the model pieces and used that contextual data to better provide instruction.

 

“We’ve seen fascinating dynamics where AI systems observe and analyze contextual data, and while we must verify the insights made by those AI agents, this creates a cycle of shared learning and adaptation that enhances the team’s performance,” Mohammad said.

 

I/ITSEC is the world’s largest modeling, simulation and training (MS&T) event that occurs every year during the week immediately after Thanksgiving. It draws thousands of MS&T professionals to the Orange County Convention Center, to include high-ranking civilian and military government officials, C-suite industry executives, and senior academics from leading research institutions.