Exploring Long-Term Human-Robot Interaction

Nowadays the first vacuum cleaning and lawn mowing robots have already entered peoples homesand daily lives. Finally, the first companies, suchas bluefrogrobotics (http://www.bluefrogrobotics.com/) and JIBO (https://www.jibo.com/) claim to produce soon available companion robots for end users. Social companion robots are intended to serve two main aspects: (i) supporting the human in useful tasks and (ii) using social cues to establish a perceived social relation to its user. BUDDY is one of the first commercially available robot claiming that it is capable of providing this type of a human-robot relationship experience. However, it is not explored how this type of robot (i.e. not focussing on a specific task to support humans in daily live, but on companionship triggered by simulated social cues) will affect social household dynamics over a longer time of period. The working hypothesis is that the interaction of ordinary users with a social companion robot) changes over time and that its adoption and acceptance differ in terms of the socio-demographic qualities of the involved households. In a series of studies in Vienna, Austria, with demographically diverse households of which each is provided with a BUDDY robot, long-term usage will be explored in order to identify factors that can promote and hinder the integration of this type of social companion robot in different types of households. The main goal is to gain detailed empirical evidence on the adoption and acceptance processes from a relatively small sample (eight households) over a long time period (eight months), therefore an ethnographic approach is chosen as methodology. Several household visits accompanied by specific field tools and techniques will be performed. A baseline of existing household routines will be established in a pre-visit before households receive a BUDDY and are firstly regularly visited over a period of six months. After half a year, the absence of BUDDY and its impact on the social household dynamics will be studied by giving it back for one month, after a one-month break. Up to now long-term studies of ordinary people sharing their space with robotic products in their homes are still rare and mainly performed with service robots, such as autonomous vacuum cleaners, but not social companion robots. Hence, there is a need for ecologically valid studies, performed outside of laboratory environments, with this novel type of robotic product. This project investigates the changing dynamics of Human-Robot Interaction in domestic environments with the social companion robot BUDDY. Comparable studies were not possible until now, as social companion robots have not achieved the technology readiness level, but were mainly studied in the laboratory or short-term in actual user homes.

What are the key dynamics and factors that influence how people perceive, understand, and ultimately adopt social companion robots in their homes? In recent years, several companies launched so-called first-wave social robots, such as Kuri, Mykie, JIBO, HUB, and Buddy, which have impacted lay people's understanding of social robots as stationary, speaking, entertainment companions. Previous research suggested that the first wave of commercially available social robots for private homes was doomed to fail. Robot start-ups shut down and participants in long-term field trials refused to use these robots after only weeks. The Vector robot, developed by Anki in 2018, is one of these first-wave social robots. It was promoted as "a robot companion and helper for people at home" and marketed for 250$. Within this Elise Richter grant, we conducted a household study in Vienna, Austria over the course of seven months, during which the Vector robot was introduced to eight demographically diverse private homes for participants' free use. The study aimed at understanding which factors of owning Vector were relevant in the long-term, even if the participants did not sustainably integrate the robot into their daily lives. Our results show the significance of subjective perceptions and expectations for what social companion robots are, how they work, and what they are or are not capable of doing. Due to a lack of perceived practical benefits and the decline of hedonic and social gains, the participating households did not sustainably integrate Vector into their daily lives. Additionally, during the duration of our study the COVID-19 pandemic led to a change in our study design. The fourth of in total six household visits happened during the first Lockdown (March 2020) and had to be conducted online, as all subsequent ones. We raised the question if being isolated at home increased (or changed) peoples' engagement with Vector (as we expected people to have more time with the robot and a higher need for entertainment and companionship). Interestingly, our interviews revealed that novel use cases emerged during the shutdown, but that the overall engagement with Vector did not increase. Based on our findings, we conclude that the adoption process of personal companion robots slightly differs from that of functional vacuum cleaning robots but may not differ substantially from technology adoption processes in general. In other words, for companion robots to succeed in future, to our conviction, we need to find a utility as a baseline to create a successful adoption. Hedonic and social gains alone do not seem to be enough to achieve a sustainable integration in people's homes.