Collaborative Intelligence and Hyperscanning: Exploring AI Application to Human-Robot Collaboration Through a Neuroscientific Approach

Cobots are robots designed to work with human operators in a shared workspace and on a shared task. Combining robots and human skills is one of the main advantages of human-robot collaboration (HRC) in industrial production. With the goal of moving toward an authentic collaboration, rather than a simple coexistence, cobots should be able to adapt to the physical and mental needs of the operator in a more natural and personalized way. Using neuroscientific measurements of human responses to HRC combined with artificial intelligence (AI) algorithms, cobots could be implemented with the ability to process and respond in real time to the psychophysiological state of the operator. Moreover, real-world scenarios must consider the presence of complex and multiple social interactions. In line with this perspective, the neuroscientific “hyperscanning” paradigm is particularly suited for the study of complex and naturalistic interactive dynamics and can be used to assess the neurophysiological activity of two or more agents interacting with each other, when a non-human agent, such as a cobot or an AI system, is introduced in the collaboration. This contribution describes a research project in early stages of development which aims to assess the effects of HRC and, more generally, of human factors, on operators’ mental and emotional state and to develop models of real-time adaptation of the cobot to the psychophysiological state of the operator.