Application Area

Human Digital Twins

We develop computational representations of human cognition, emotion, decision-making, and behavior. These dynamic models combine multimodal data and artificial intelligence to estimate current human states, anticipate future responses, and support personalized interventions.

Research Vision

Computational Models of the Individual

A human digital twin is more than a static profile. It is an evolving computational model that learns from physiological signals, behavioral observations, interaction histories, environmental conditions, and contextual information.

By continuously updating this representation, intelligent systems can better understand individual differences, changing needs, and likely future responses.

Core Capabilities

Understand, Predict, and Personalize

Our research investigates how digital twins can represent complex human states and support adaptive decisions across healthcare, education, training, and interactive systems.

Cognitive Modeling Emotion Modeling Behavior Prediction Personalization Multimodal AI Decision Support

How It Works

From Human Signals to an Adaptive Digital Representation

Human digital twins are created through a continuous cycle of sensing, modeling, prediction, and adaptation.

01 · Observe

Capture Multimodal Human Data

Collect information from EEG, eye tracking, physiological sensors, movement, task performance, interaction behavior, and environmental context.

02 · Model

Build a Dynamic Human Representation

Use machine learning and computational models to estimate cognitive, emotional, behavioral, and performance-related states.

03 · Adapt

Predict and Support Future Decisions

Forecast likely responses, identify emerging risks, and personalize interventions, environments, or recommendations in real time.

Application Domains

Where Human Digital Twins Can Make an Impact

Healthcare Intelligence

Personalized Monitoring and Intervention

Digital twins can support individualized health monitoring, rehabilitation planning, mental-health assessment, treatment personalization, and early identification of changing patient needs.

Training and Simulation

Adaptive Training for Individual Performance

Training systems can adjust task difficulty, feedback, and scenario complexity according to workload, stress, fatigue, expertise, and learning progress.

Behavior Prediction

Anticipate Decisions and Performance

Computational twins can help estimate future choices, performance changes, cognitive overload, disengagement, or risk under different environmental conditions.

Human-Centered Systems

Create Technology That Learns the User

Intelligent interfaces, autonomous systems, and immersive environments can use digital-twin models to provide more meaningful, transparent, and personalized interactions.

Research Challenges

Building Digital Twins That Are Accurate, Ethical, and Trustworthy

Human behavior is dynamic, context-dependent, and difficult to capture with a single model. Our research therefore addresses uncertainty, personalization, temporal change, limited data, interpretability, fairness, privacy, and responsible use.

Explainable AI Privacy Fairness Uncertainty Longitudinal Modeling Responsible AI

Student Opportunities

Build Intelligent Models of Human Cognition and Behavior

Students can contribute to multimodal data collection, physiological signal analysis, behavioral modeling, time-series prediction, personalized machine learning, simulation, explainable AI, and human-centered evaluation.

Projects may involve developing digital-twin prototypes, predicting cognitive or emotional states, designing adaptive interventions, or studying how computational models can safely support real-world decisions.

Python Machine Learning Time-Series Modeling Signal Processing Simulation Human Studies