Pusceddu Giulia, Cocchella Francesca, Bogliolo Michela, Belgiovine Giulia, Lastrico Linda, Rea Francesco, Casadio Maura, Sciutti Alessandra
COgNiTive Architecture for Collaborative Technologies, Italian Institute of Technology, Genoa, Italy.
Department of Informatics, Bioengineering, Robotics and Systems Engineering, University of Genoa, Genoa, Italy.
Front Robot AI. 2025 Apr 1;12:1563923. doi: 10.3389/frobt.2025.1563923. eCollection 2025.
This work aims to advance the understanding of group dynamics in robot-child interactions, focusing on whether, during a motor-imitation task led by a Nao robot, children might be influenced in their action executions by other group members - human or robotic. After testing eighteen groups of four children and teenagers, our findings indicate that participants tend to disregard the robot when it performs atypical gestures, preferring instead to imitate the actions of a human peer. Moreover, we found evidence that, in this scenario, assigning a leadership role to the robot does not, by itself, guarantee compliance from human group members; broader group dynamics must also be taken into account. Further results show that participants are significantly more likely to imitate the robot's action when the "proactive" group members (i.e., those who initiate actions first) conform to Nao, compared to when they do not. Previous studies suggest that the mutual influence of group members can facilitate interaction with a robotic agent; however, our findings show that the presence of proactive members could also undermine the group's conformity to the robot. Additionally, these findings highlight the importance of personalizing robots to better integrate into specific group dynamics, enhancing their ability to influence different groups effectively.
这项工作旨在增进对机器人与儿童互动中群体动态的理解,重点关注在由Nao机器人主导的动作模仿任务中,儿童在动作执行过程中是否会受到其他群体成员(人类或机器人)的影响。在对18组由四名儿童和青少年组成的群体进行测试后,我们的研究结果表明,当机器人做出非典型手势时,参与者往往会忽视它,而是更倾向于模仿人类同伴的动作。此外,我们发现有证据表明,在这种情况下,仅给机器人赋予领导角色并不能保证人类群体成员的服从;还必须考虑更广泛的群体动态。进一步的结果表明,与“主动”群体成员(即那些首先发起动作的成员)不遵循Nao的情况相比,当他们遵循Nao时,参与者模仿机器人动作的可能性要大得多。先前的研究表明,群体成员之间的相互影响可以促进与机器人代理的互动;然而,我们的研究结果表明,主动成员的存在也可能破坏群体对机器人的顺从。此外,这些研究结果凸显了使机器人个性化以更好地融入特定群体动态的重要性,从而增强它们有效影响不同群体的能力。