Understanding how two people coordinate while handling the same object remains one of the most elusive challenges in manual assembly. When accuracy and timing must align perfectly, as with placing a car windshield, even minor miscommunication can lead to misalignment, quality issues or safety risks. Yet capturing this coordination reliably has proved difficult due to drift, poor synchronisation, and the limitations of conventional motion capture systems.

Mapping collaboration with precision

This work presents an experimental setup designed to capture high-quality motion data from two humans jointly manipulating an object. Combining full-body inertial tracking, finger sensing and gaze recording, the system captures human movement in detail while maintaining synchronisation across devices. An overhead projector provides visual cues for initial and target positions, ensuring consistent task execution across repeated trials.

A digital twin environment supports the setup by enabling feasibility checks before physical trials. This allows early identification of errors and ensures that target positions remain within the reachable area of both participants.

Reproducible insights into shared movement

The setup records hand trajectories, object movement and eye focus while two participants perform repeated pick-and-place tasks. Early results show stable spatial coordination, identifiable leader-follower roles and clear motion patterns that can be analysed for timing, deviation and synchronisation. These findings provide a basis for modelling collaborative behaviour and understanding how humans communicate through movement alone.

Towards human-aware robotic collaboration

The system’s reproducibility makes it a powerful tool for developing digital human models and future human-robot collaboration strategies. High-fidelity motion capture allows researchers to extract latent motion features that can later support robot control, enabling robots to understand and adapt to human intentions.

This article is based on the peer-reviewed publication “Experimental setup for motion capturing two humans jointly handling an object”, published in Procedia CIRP (Volume 134, 2025).