Human-Human Connected Dyads Learning a Visuomotor Rotation in a Targeted Reaching Task.

Little is known about how two people physically coupled together (a dyad) can accomplish tasks. In a pilot study we tested how healthy inexperienced and experienced dyads learn to repeatedly reach to a target and stop while challenged with a 30 degree visuomotor rotation. We employed the Pantograph investigational device that haptically couples partners movements while providing cursor feedback, and we measured the amount and speed of learning to test a prevailing hypothesis: dyads with no experience learn faster than an experienced person coupled with a novice. We found significant straightening of movements for dyads in terms of amount of learning (2.662±0.102 cm and 2.576±0.024 cm for the novice-novice and novice-experienced groups) at rapid rates (time constants of 17.83 ± 2.85 and 18.17.17±6.72 movements), which was nearly half the learning time as solo individuals' studies. However, we found no differences between the novice-novice and experienced-novice groups, though retrospectively our power was only 3 percent. This pilot study demonstrates new opportunities to investigate the advantages of partner-facilitated learning with solely haptic communication which and can lead to insights on control in human physical interactions and can guide the design of future human-robot-human interaction systems.

Human-Human connected dyads learning a visuomotor rotation in a movement tracking task.

Dyads are couples of collaborative humans that perform a task together while mechanically connected by a robot. As shown in different studies [1] [2], haptic interaction can be beneficial for motor performance so that the dyad outperforms the subject executing the task alone. These achievements are hypothesized to be the result of the haptic communication engaged between the subjects that triggers internal forward models. In this way the dyad's components can attain additional information about the task, hence improving their performance. Here we show a novel dual robotic system, called Pantograph, used in a pilot study to understand the influence that the nature of the partner has on the learning process. The main hypothesis that we claim is that a Novice-Novice type of interaction is more beneficial, in terms of speed of learning, with respect to an Expert-Novice type of interaction. The results show time constants equal to 5.53 ± 2.79 and 8.45 ± 3.78 for the Novice-Novice and Expert-Novice group, respectively. However, the p-value obtained was p = 7.54%. Hence, we can not generalize our results, but this research study shows how haptic communication between interacting humans allows for motor learning and how the nature of the subjects could be an important factor of the learning process.