Title: Supporting Healthy Dyadic Human Relationships With Power Differentials Using Robots



Michael Pettinati


Ph.D. Student


School of Interactive Computing


College of Computing


Georgia Institute of Technology



Date: Tuesday, August 1, 2017

Time: 4:00-6:00pm (EST)


Location: TSRB 223



Committee:
------------------------


Dr. Ronald Arkin (Advisor, School of Interactive Computing, Georgia Institute of Technology)


Dr. Rosa Arriaga (School of Interactive Computing, Georgia Institute of Technology)


Dr. Sonia Chernova (School of Interactive Computing, Georgia Institute of Technology)


Dr. Ayanna Howard (School of Electrical and Computer Engineering, Georgia Institute of Technology)

Dr. Mark Riedl (School of Interactive Computing, Georgia Institute of Technology)



Abstract
------------------------


Satisfying relationships are not relationships that lack conflict; they are relationships where conflict is managed in healthy ways. Dyadic relationships with power differentials often see one dyad member make a choice to use an unhealthy conflict strategy. An agent that facilitates the use of healthier interaction styles would be invaluable in these moments for these relationships. This proposal explores how a robot can fulfill this role and help human dyads attain and sustain mutually satisfactory relationships. 

The proposal presents preliminary results that suggest robots may have certain qualities, such as their more limited social presence, that make them desirable for this role. It introduces a computational model that identifies states of dissatisfaction in dyadic human relationships with power differentials. It also proposes an action-selection mechanism that selects interventions to help ameliorate states of dissatisfaction. Finally, the proposal gives an overview of a two-phase human-robot interaction experiment to test an implementation of this computational architecture and provide more definitive answers regarding how a robot can support healthy dyadic interactions in relationships with power differentials.