Data from: Robot-mediated haptic feedback outperforms vision in violin duo coordination

Joint actions among humans rely on the integration of multiple sensory
modalities, most notably auditory and visual cues, which support explicit
communication between partners. However, haptic feedback provides a
direct, implicit channel for sensorimotor communication, and its
contribution to fine motor coordination in joint actions remains largely
unexplored. Here, we demonstrate that haptic communication, rendered
through bidirectionally coupled wearable robots, outperforms traditional
auditory-visual feedback in a complex and challenging real-life joint
action: ensemble violin performance. First, we developed a pair of
2-degree-of-freedom upper-limb exoskeletons capable of transparently
following violinists’ natural movements and rendering viscoelastic torques
proportional to the joint angular deviation between the partners. Then, we
designed a within-subject experiment with twenty violin duos performing a
musical piece under four sensory feedback conditions: auditory (A),
auditory-visual (AV), auditory-haptic (AH), and auditory-visual-haptic
(AVH), across two tempi (72, 100 bpm). Despite the musicians being
unfamiliar with the robot-mediated haptic feedback and unaware of the
bidirectional connection between them, haptic feedback (AH, AVH)
significantly enhanced spatiotemporal coordination and dynamic musical
alignment compared to the extensively trained auditory-visual feedback (A,
AV). Notably, the multisensory feedback condition AVH yielded the highest
scores across all measures. Our findings demonstrate that haptic feedback
can support fine motor coordination in violin duo performance more
effectively than visual cues, particularly for professional musicians, due
to its implicit and embodied nature, and that it can be effectively
delivered via wearable robots, paving the way for new paradigms in
human-human sensorimotor interactions.