Studying Effect of Expressive Light Animation on Human Perception and Behavior of Human-Machine Interaction

Sichao Song1, Seiji Yamada2

  • 1The Graduate University for Advanced Studies (SOKENDAI)
  • 2National Institute Informatics

Details

10:00 - 10:30 | Mon 25 Sep | Ballroom Foyer | MoAmPo.45

Session: Monday Posters AM

Abstract

Although many studies have investigated human-like interaction methods involving facial expressions and gestures, there has been little focus on appearance-constrained robots due to their restrictions in interaction modalities. Existing approaches for promoting affective interaction for such robots mainly focus on motion cues or body posture. However, such approaches lack expressibility and are hard, if not impossible, to apply in many application scenarios. To address this issue, we investigate an alternative modality, expressive light, as an affective interaction modality. Compared with other modalities such as motion cues, body posture, and acoustic cues, expressive light offers a much broader design space, is less intrusive, and is less restricted to spatiotemporal limitations. However, until now, it has been unclear how robots can take advantage of expressive light as a primary modality to express affect, and furthermore, affect people’s perceptual behavior toward robots. To investigate this, we performed two experiments using an Ultimatum game and a Give-Some game in which we attached an LED strip to the front-bottom of a computer monitor and had it display the two light animations, green and low-intensity (GL) and red and high-intensity (RH). We show that our results are in line with color psychology theories. Specifically, the participants perceived the GL light animation positively and further displayed approach-like behaviors by showing a higher tolerance regarding unfair offers and behaving more cooperatively. Similarly, they perceived the RH light animation negatively and further displayed avoidance-like behaviors by showing a lower tolerance regarding unfair offers and behaving less cooperatively (selfishly). In addition, an analysis of a post-experiment questionnaire showed that the participants used positive adjectives such as “friendly” and “kind” to describe the computer when it displayed the GL light animation and negative adjectives such as “angry” and “oppressive” when it displayed the RH light animation. We are now working on installing an LED lighting system in an iRobot Create 2 robot and are going to use it as a robotic platform to translate this study to human-robot interaction scenarios.