Introduction. Recent developments in Cognitive Human-Robot Interaction (CHRI) have shifted the role of collaborative robots (cobots) from simple automated machines to intelligent partners able to interpret human intentions and adapt accordingly. In manufacturing, this transformation enables new interaction paradigms such as dynamic task allocation, where task assignments can be redistributed during work tasks according to user needs. While such dynamic adaptations should rely on equally dynamic communicative exchanges between the robot and the human, existing systems predominantly envision unidirectional communication—either cobot-to-human or human-to-cobot.


Purpose. With this work, we move beyond the robot-centric perspective and compare different human-robot communication protocols, including bidirectional human-cobot (BUC), unidirectional cobot-to-human (U-CU), and unidirectional human-to-cobot (U-UC) communication. We specifically focus on human perspectives and address the system's implications for psychological well-being.


Methods. We run an experimental study involving 34 participants collaborating on an assembly task involving dynamic task allocation in a shared workstation. The experiment took place in our highly ecological lab environment simulating an industrial workcell, where all participants completed the assembly task under three communication conditions (i.e., BUC, U-CU, U-UC). Data collected included explicit and implicit measures of cognitive load gained through self-reports and eye-tracking, perceived and behavioral performance metrics, and self-reported assessments of communication quality and teamwork.


Results. Results show that although BUC required longer communicative exchanges, participants rated it as the least frustrating condition. Task performance remained comparable across communication modalities, while BUC demonstrated the highest scores for perceived teamwork and communication quality. Interestingly, participants in the BUC condition evaluated identical allocation suggestions more positively than U-CU, highlighting the role of reciprocal communication in shaping collaboration quality.


Conclusions. These preliminary findings suggest that human-robot communication systems envisioning bidirectional communication may promote a human-cobot synergy that sustains performance without losing focus on worker satisfaction and psychological well-being.