Abstract:
The integration of small Unmanned Aircraft Systems (sUAS) into low-altitude urban airspace is gaining momentum. Safety challenges arise, necessitating automated (and in s...Show MoreMetadata
Abstract:
The integration of small Unmanned Aircraft Systems (sUAS) into low-altitude urban airspace is gaining momentum. Safety challenges arise, necessitating automated (and in some cases autonomous) technical solutions. In order for the operator to monitor and engage with the sUAS an effective human-machine interface (HMI) is a vital component. Equally, the manner in how information is presented on this HMI requires careful consideration – more so when the operator may very well have more than one aerial platform under their control and/or supervision. This study explores the role of icon design for autonomous sUAS supervisory control to contribute to the growing body of research on explainable artificial intelligence. Altogether, 14 icons are proposed as representations of safety-critical functions related to autonomous sUAS operation in low-altitude urban airspace. In an online questionnaire study, 46 participants with experience in operating sUAS rated the icons on established icon-function fit metrics. The analysis of agreement scores indicates that the icons related to battery health, geofence conformance, and meteorological constraints were well-recognized, while those representing casualty risk, positional accuracy, airspace conformance, and sensor health performed poorly. These findings emphasize the importance of concreteness, familiarity, and semantic distance in icon design, where higher values positively influence icon recognition and thus icon-function fit. The integration of empirically derived icon design principles is proposed to enhance transparency in safety-critical autonomous systems. This study underscores the significance of targeted usage of unambiguous icons to facilitate deeper user understanding through making system-side decision-making processes transparent to the user, enabling more effective interaction between humans and autonomous systems.
Date of Conference: 15-17 May 2024
Date Added to IEEE Xplore: 19 June 2024
ISBN Information: