Research

Research at HILS

Our research connects human intelligence, artificial intelligence, and robotics across five complementary directions.

Neuroergonomics & Cognitive State Modeling

We use neurophysiological and behavioral sensing to model cognitive states that shape human interaction with complex systems. Our work examines workload, attention, situational awareness, trust, reaction time, motor-control difficulty, and engagement using EEG, EMG, gaze, and other physiological measurements.

Brain-computer interfaces and neuroergonomics
Cognitive workload, attention, trust, and situational awareness
Neurophysiological modeling for human–swarm and human–robot interaction
Multimodal classification of reaction time and motor-control difficulty

AI for Human Health & Performance Modeling

We develop AI-enabled digital models of human health and performance that combine physiology, behavior, context, and environment. Applications range from rehabilitation and mobility monitoring to fatigue assessment, hiking-time prediction, training, and human performance optimization.

State-dependent hiking and traversal-time prediction
Physiological and wearable-sensor models of fatigue and gait
Home-based rehabilitation and participation monitoring
Foundation models for nonstationary physiological time series

Physiological time-series foundation model architecture

Current directionFoundation models for physiological time series using stochastic intrinsic-mode representations and adaptive multiresolution learning.

Human-Inspired AI & Human-in-the-Loop Learning

Human information processing can provide powerful inductive biases for AI. We develop learning approaches that incorporate gaze, eye movement, physiological feedback, and context-dependent human behavior to improve representation learning, imitation learning, reinforcement learning, and adaptive assistance.

Gaze-guided contrastive representation learning
Selective eye-gaze augmentation for imitation learning
Human context-aware action modification for AI assistance
Physiological feedback in adaptive learning systems

Human–AI & Human–Robot Teaming

We study how people collaborate with autonomous agents and robots, from multi-robot supervision and human–swarm interaction to direct physical collaboration. The goal is to make teaming safer, more adaptive, more understandable, and better matched to human capabilities and state.

Human–swarm interaction and adaptive autonomy
Human–UGV cooperation and casualty evacuation
Physical human–robot interaction and admittance control
Trust, workload, tactical decisions, and shared control

Robotic Design & Industrial Automation

We design robotic systems and mechanisms that can operate safely and effectively in uncertain physical environments. Current work includes compliant manipulation, variable stiffness, tactile sensing, robot–soft-body interaction, sim-to-real transfer, industrial automation, and advanced biomanufacturing.

Variable-stiffness and compliant end-of-arm tooling
Dexterous tactile sensing and contact-rich manipulation
Robot–soft-body interaction and sim-to-real transfer
Industrial automation, packaging, and biomanufacturing

Current directionSim-to-real transfer and contact-aware robotic manipulation for soft and deformable objects.

Collaborate

Interested in working with HILS?

We welcome interdisciplinary collaboration across AI, robotics, human performance, neuroergonomics, and health.

Contact the Lab