This February, an interactive, collaborative workshop between IEEE Systems, Man, and Cybernetics Society (SMCS) and the Institute of Simulation and Training, School of Modeling, Simulation, and Training at the University of Central Florida was held to explore the intersection of Human-Machine Systems (HMS) and Digital Twin Technologies (DTT).
The workshop, chaired by Dr. Soheil Sabri (UCF) and Dr. Saeid Nahavandi (IEEE), hosted attendees and presenters from across the globe, dove into the theory, emerging technologies, and real-world applications of DTT in HMS applications. In addition to the scope of these applications, the difference in how each organization viewed, utilized, and defined DTT and Artificial Intelligence (AI). Carolina Cruz-Neira, Agere Chair Professor in the Department of Computer Science at the University of Central Florida demonstrated how DTT can be used to more efficiently demonstrate data. According to Cruz-Neira, we don’t use more than 20% of the data we collect, and when we relegate data to dashboards, it is pulled out of its context and still requires human interpretation. Using DTT, users are able to put data in its original context as well as introducing a level of interactivity with data that is not present on a dashboard.
However, DTT is not only useful as a means to decipher data, but also to create it. Tadahiko Murata, Professor at the Cybermedia Center of Osaka University and IEEE Fellow, explained how Japan is using DTT to create a synthetic population and household projections. This technology has created simulations to predict voting patterns, the labor market, and even the spread of Covid-19. This is made possible by creating a digital twin based on census data, developing a synthesized population of every prefecture, breaking it down into cities, towns, villages, and even the “small areas” that make up those populations.
With 47 prefectures in Japan, gathering factual data on every person within the population is impossible. However, Murata explains that every family in Japan has been synthesized in this way. Utilizing factual data and AI, this synthesized population is able to react in a fashion that predicts real-world outcomes. The simulation is run multiple times and the likelihood of outcomes is determined via a bell-curve. The decisions made by this synthesized population may not directly reflect the decisions made by their real-world counterparts, but it is able to predict the decisions of a larger population.
AI systems are at the core of DTT, being utilized to inform the decisions made by “human” digital twins (HDT). However, one of the pitfalls of AI that was discussed at the workshop is security. Whether you are utilizing AI to function as a digital financier or as a digital girlfriend, the codebase remains the same. The largest number of HDTs is found in the adult entertainment industry, seen in models like Camryn AI. According to Ben D. Sawyer, Associate Professor at the University of Central Florida, these “adult” digital twins are a “security nightmare,” being used to leverage as much data as possible before legislation catches up.
According to Dr. Joseph Cohn of SoarTech, “until we get a grasp on how we can best use AI systems, we’ll find ourselves in increasingly strange social situations where decisions are made without a full understanding of how humans and AI can work together.” It is still being determined when it is appropriate to create a digital twin versus a digital clone, or when artificial thinking is fallible. DTT is evolving, being defined and redefined by the very people who presented at this workshop.
This workshop was intended to increase awareness of real-world applications and case studies of successful DTT, focusing on HMS and collaboration, and to deepen the understanding of these technologies. It was an incredibly enlightening event, lending practical insights into integrating these emerging technologies for improved system performance.
IEEE will continue this conversation in May at their International Defence Excellence and Security Symposium (IDEaS) in Toronto, Canada. In addition, the IEEE SMC will publish the recordings of this workshop on their website, early next month.

People who read this article also found these articles interesting :