A Systematic Review of Trust Assessments in Human–Robot Interaction

摘要

The integration of robots into daily life has increased significantly, spanning applications from social-care to industrial settings with collaborative robots. Ensuring a safe, secure environment and equitable workload distribution in human-robot collaborations is crucial. Trust is a key factor in these environments, essential for enhancing collaboration and achieving tasks while maintaining safety. Under-trusting robots can hinder productivity, while over-trusting them can lead to accidents. A comprehensive literature review of 100 publications from 2003 to 2023 analyzes trust and its influencing factors in industrial and social-care contexts. Findings reveal that in industrial settings, robot-related factors are more influential, whereas in social-care, human and environmental factors play a significant role. Furthermore, the review delves into gauging trust through observable behavior, while also exploring various trust evaluation methodologies. Results show that trust can be gauged through human behaviors, physical cues, and physiological measurements. Concerning trust evaluation methodologies, traditional questionnaires have limitations, opening new opportunities for machine learning and sensor-based approaches to real-time trust evaluation, as trust is a dynamic cognitive value that evolves over time. Notably, 97% of the reviewed articles were published in the last decade, underscoring a growing interest in human–robot interaction and trust within the scientific community.