Emotional Expressions of Real Humanoid Robots and Their Influence on Human Decision-Making in a Finite Iterated Prisoner’s Dilemma Game

Abstract

Emotional expressiveness in robots has attracted increased research interest, and development of this technology has progressed in recent years. A wide variety of methods employing facial expressions, speech, body movements, and colors indicating emotional expression have been proposed. Although previous studies explored how emotional expressions are recognized and affect human behavior, cooperative and competitive human-robot relationships have not been well studied. In some cases, researchers have examined how cooperative relationships can be influenced by agents that exhibit detailed facial expressions on a screen. It remains unclear whether and how expressive whole-body movements of real humanoid robots influence cooperative decision-making. To explore this, we conducted an experiment in which participants played a finite iterated prisoner’s dilemma game with a small humanoid robot that exhibits multimodal emotional expressions through limb motions, LED lights, and speech. Results showed that participants were more cooperative when the humanoid robot showed emotional expression. This implies that real humanoid robots that lack ability to show sophisticated facial expressions can form cooperative relationships with humans by using whole-body motion, colors, and speech.