Mateusz Woźniak
Italian Institute of Technology, Genoa, Italy
Being an agent involves having at least some degree of agency or control over one’s actions. The sense of agency is the subjective/experiential aspect of having agency and has been regarded as one of the fundamental aspects of one’s sense of self (Gallagher, 2000). Moreover, the sense of agency also serves as an important part of a mechanism through which we can incorporate objects and entities into our self-representation. For example, experiencing that I have agency over a digital avatar or a robot might lead me to represent it as part of my extended self: my digital or robotic body. However, while having complete control over one’s avatar might maximally facilitate sense of agency and self-identification with the avatar, practical concerns often motivate us to implement various levels and types of autonomy into our avatars. As a consequence, our avatars can end up being entities that are either closer to being our external bodies, or to being partners of interaction. In this talk I will provide an overview of our recent research on psychological and neurocognitive processes that are involved and affected when one controls a semi-autonomous robot using different forms of shared control architectures. I will summarize our studies on how shared control in human-robot interaction can influence sense of agency, trust, sense of joint agency, as well as subjective and objective performance.
