AUTHOR=Gul Aisha , Turner Liam , Fuentes Carolina 

TITLE=Conventions and research challenges in considering trust with socially assistive robots for older adults

JOURNAL=Frontiers in Robotics and AI

VOLUME=Volume 12 - 2025

YEAR=2025

URL=https://www.frontiersin.org/journals/robotics-and-ai/articles/10.3389/frobt.2025.1631206

DOI=10.3389/frobt.2025.1631206

ISSN=2296-9144

ABSTRACT=IntroductionThe global ageing population rise creates a growing need for assistance and Socially Assistive robots (SARs) have the potential to support independence for older adults. However, to allow older adults to benefit from robots that will assist in daily life, it is important to better understand the role of trust in SARs.MethodWe present a Systematic Literature Review (SLR) aiming to identify the models, methods, and research settings used for measuring trust in SARs with older adults as population and analyse current factors in trust assessment.ResultOur results reveal that previous studies were mostly conducted in lab settings and used subjective self-report measures like questionnaires, interviews, and surveys to measure the trust of older adults in SARs. Moreover, many of these studies focus on healthy older adults without age-related disabilities. We also examine different human-robot trust models that influence trust, and we discuss the lack of standardisation in the measurement of trust among older people in SARs.DiscussionTo address the standardisation gap, we developed a conceptual framework, Subjective Objective Trust Assessment HRI (SOTA-HRI), that incorporates subjective and objective measures to comprehensively evaluate trust in human-robot inter-actions. By combining these dimensions, our proposed framework provides a foundation for future research to design tailored interventions, enhance interaction quality, and ensure reliable trust assessment methods in this domain. Finally, we highlight key areas for future research, such as considering demographic sensitivity in trust-building strategies and further exploring contextual factors such as predictability and dependability that have not been thoroughly explored.