The contribution of people’s attitudes and perceptions to the acceptance of eldercare robots

Abstract

The number of people aged over 65 years is increasing worldwide, and this is placing increased demand on healthcare services. Engineers have proposed that eldercare robots may be able to meet the increasing healthcare needs of the aging population; however eldercare robots have not yet been widely adopted. Reasons for this are likely multifaceted, but one reason may be insufficient attention to the psychological aspects of the human robot interaction (HRI) in eldercare. Technology acceptance models indicate that people’s perceptions of technology attributes (particularly perceived usefulness) predict technology acceptance more strongly than more objective design parameters. However, little research to date has investigated the importance of perceptions to the acceptance of eldercare robots. The central thesis of this PhD is that older people’s perceptions will influence their acceptance of healthcare robots. Specifically, three main perceptions are studied - older people’s perceptions of their own unmet needs, their attitudes towards robots in general, and their perceptions of the robot’s mind. It is proposed that more positive attitudes and perceptions of robots will predict better acceptance of healthcare robots. This thesis contains four peer reviewed publications. One is a discussion paper on the importance of assessing the unmet needs of eldercare stakeholders in order to develop more useful and acceptable robots. Three publications present the results of three different Human Robot Interaction (HRI) trials conducted with prototype healthcare robots. All three studies employed autonomous service-type robots and older participants, and two of the three HRI trials were conducted within real-world eldercare environments. The key findings of the HRI studies were that people’s perceptions of robots and ‘robot mind’ predicted robot acceptance. In all three studies, participants’ ‘pre-interaction’ generic robot attitudes predicted acceptance of specific robots. This suggests that even people who have never used robots before can hold mental models of robots that influence robot acceptance. Additionally, people’s robot attitudes improved after interacting with the robot, and these changes also predicted robot acceptance. This suggests that a positive HRI is important for robot acceptance. Compared with people who perceived robots as possessing more mind, people who perceived robots as having less mind were more likely to use a robot. Furthermore, despite robot-users perceiving less robot-mind at baseline, they perceived the robot to have even less mind after interacting with it. While this result suggests that people may hold unrealistically high perceptions of a robot’s mind which may be a barrier to acceptance, it also suggests that these perceptions are revised downwards after actually experiencing a robot’s capabilities. In conclusion, older people’s perceptions and attitudes towards robots do predict eldercare robot acceptance. Future implications of this work are that building robots that meet the specific unmet needs of older people and paying more attention to users’ perceptions of robots may increase the acceptance of eldercare robots. Future research should investigate whether interventions designed to promote realistic and adaptive perceptions of robots in older people can increase the acceptance of eldercare robots.