Are we working on the safety of UAVs? An LDA-based study of UAV safety technology trends

Abstract

This study aims to investigate whether the Unmanned Aerial Vehicle (UAV) safety studies match the safety issues relating to the UAV goods. The causes and the effects of the research status are combined to predict the future development of the field. To identify core categories and core concepts of the UAV safety issues, this study conducts qualitative research on the bad reviews of the UAV goods sold on e-commerce platforms. By using the Latent Dirichlet Allocation (LDA) model, the topics of the research papers on the Chinese National Knowledge Infrastructure (CNKI) and the Web of Science (WoS) are derived. The findings show that there is a gap between the UAV studies and the safety issues relating to UAV goods. Furthermore, a majority of UAV safety studies are focused on information transfer and algorithm optimization. In comparison, the studies on safety improvement from the hardware perspective are not prominent. Accordingly, this study argues that UAV development should be coordinated by regulation, supervision, and even price leverage. The paths, both data-information-network and sensing-control-optimization, could be of interest to future studies on UAV safety technologies.