Evaluation Method for Robot Interfaces Considering Psychological Safety

Abstract

Service robots have been implemented across various sectors in recent years to enhance work efficiency. Developing interfaces with an emphasis on usability from a human-centered design perspective is crucial. Numerous evaluation methods have been suggested for robot interfaces, including those focusing on work efficiency and error rate reduction. Although the NASA-TLX safety evaluation method can assess specific indicators in hazardous situations, an evaluation methodology that emphasizes psychological safety is required. In this study, we propose an evaluation approach that considers psychological safety, utilizing a differential two-wheeled robot. Furthermore, we conducted experiments using three different interfaces. The result of empirical experiments elucidated that interfaces characterized by elevated psychological safety engendered a proclivity among participants towards diverse exploratory methodologies and the adoption of venturesome undertakings. Conversely, interfaces marked by diminished psychological safety precipitated impediments for participants, largely attributed to the palpable absence of tactile reinforcement, thereby obfuscating the stimulation of audacious endeavors. Such revelations accentuate the paramountcy of cognizing the nuanced gradations in psychological safety that modulate the propensity for immersion in exigent tasks. Psychological safety pertains to an individual's perception of the consequences of engaging in interpersonal risk-taking. Thus, it is necessary to explore psychological safety assessment models for various interfaces and conduct surveys among study participants to develop a highly accurate model. In the future, we should also consider control models for automobiles and other vehicles.