Principal Investigator: Michael Beigl

Virtual meetings have become essential in hybrid and remote collaboration, yet they often lack the subtle non-verbal and contextual signals that support coordination, emotional understanding, and a sense of group presence. This project explores how haptic feedback delivered through wearable devices can augment social awareness during video conferences. Framed within the broader theme of human-system co-adaptation, it examines how individuals and adaptive systems can iteratively adjust to one another, both at the level of moment-to-moment interaction and over the course of sustained use.

The central goal is to develop a co-adaptive haptic system that supports richer interpersonal communication in digital meetings. By translating key contextual and affective cues into real-time tactile signals, the system aims to reduce video conferencing fatigue, enhance social presence, and improve team satisfaction and performance. Personalization and adaptive feedback are core to this approach: users can customize how and when they receive tactile feedback, while the system itself learns from behavioral patterns and user states to adjust its output dynamically.

The system will be designed to integrate a range of input sources to infer socially and contextually relevant moments during a meeting, including physiological signals such as heart rate, audiovisual emotion recognition, attention estimation via eye tracking, and conversation analysis using large language model (LLM)-based chat processing. These inputs guide the delivery of haptic feedback through smartwatches, enabling participants to receive non-disruptive cues related to events such as participant reactions, conversational shifts, or attention lapses. A user study compares different interaction modalities—ranging from speech-only to visual and haptic-augmented configurations—evaluating their impact on social connectedness, perceived workload, task performance, and communication quality. Aligned with the co-adaptation framework, the system supports both user-driven personalization and system-driven learning, creating a feedback loop through which technology and users adapt to one another over time.