Social Robots on a Global Stage: Establishing a Role for Culture During Human–Robot Interaction

Abstract

Robotic agents designed to assist people across a variety of social and service settings are becoming increasingly prevalent across the world. Here we synthesise two decades of empirical evidence from human–robot interaction (HRI) research to focus on cultural influences on expectations towards and responses to social robots, as well as the utility of robots displaying culturally specific social cues for improving human engagement. Findings suggest complex and intricate relationships between culture and human cognition in the context of HRI. The studies reviewed here transcend the often-studied and prototypical east–west dichotomy of cultures, and explore how people’s perceptions of robots are informed by their national culture as well as their experiences with robots. Many of the findings presented in this review raise intriguing questions concerning future directions for robotics designers and cultural psychologists, in terms of conceptualising and delivering culturally sensitive robots. We point out that such development is currently limited by heterogenous methods and low statistical power, which contribute to a concerning lack of generalisability. We also propose several avenues through which future work may begin to address these shortcomings. In sum, we highlight the critical role of culture in mediating efforts to develop robots aligned with human users’ cultural backgrounds, and argue for further research into the role of culturally-informed robotic development in facilitating human–robot interaction.

Mind Meets Machine: Towards a Cognitive Science of Human–Machine Interactions

“…Indeed, the individual features being evaluated might be nonlinear and multidimensional themselves, with their importance varying depending on one's research questions, as well as complex contextual factors, such as how a machine is introduced to a person, the person's age, cultural background, or what they believe about that particular machine (cf. [26][27][28][29]). Finally, the dimensions outlined in Figure 1B are simple examples to illustrate the value of a dimensional (compared with a categorical) approach rather than a suggestion that these dimensions should receive privileged status.…”

“…People's general lack of experience with robots in the real world, which can be further exacerbated by cultural and socioeconomic factors (e.g., [29]), also means that the stage of learning about these machines is different from that regarding people and common objects. With such 'immature' representations (in terms of both phylogeny and ontogeny) comes more opportunity for trial-and-error learning rather than consolidated forms of learning.…”

Expect the Unexpected: Leveraging the Human-Robot Ecosystem to Handle Unexpected Robot Failures

“…Design approachable robots that provide people with advice, feedback and instructions in a manner that encourages cooperation and compliance, particularly during first encounters (e.g., by adhering to social and cultural norms). Yamamoto et al (1992), Torrey (2009), Wang et al (2010), Torrey et al (2013), Trovato et al (2013, Salem et al (2014), Geiskkovitch et al (2016, Gehle et al (2017), Choi et al (2019), Xu (2019), Hashemian et al (2019), Sanoubari et al (2019), Lim et al (2020), Avelino et al (2021), Ullrich et al (2021), Fischer et al (2021, Tian and Oviatt (2021) Description: Units should be actively encouraged to take initiative in helping other units resolve problems. Motivating mechanisms and incentives must be formed.…”

From Humans and Back: a Survey on Using Machine Learning to both Socially Perceive Humans and Explain to Them Robot Behaviours

“…While the influences of culture on expectation and responses to robot have been studied in HRI [68], it is less common in EB-specific studies. For example in Johnson et al [69], all participants were from the Netherlands, thereby increasing the chances of representing a common culture.…”