UAV Operations Safety Assessment: A Systematic Literature Review
Omid Asghari, Naghmeh Ivaki, Henrique Madeira
Abstract
The significant increase in urban Unmanned Aerial Vehicles (UAVs), due to their benefits and commercial potential, will increase drone density and collision risks. To manage this, Unmanned Aircraft Systems Traffic Management (UTM), European implementation of UTM (U-space), and Air Traffic Management (ATM) are being developed for safe integration with other air traffic. Nonetheless, thorough safety assessments remain essential for ensuring UAV operation safety. In this study, we conducted a two-phase systematic literature review. First, we analyzed existing reviews on UAV operation safety assessments. Second, we examined primary studies with the goal of identifying (i) safety assessment approaches, (ii) employed methods/techniques, (iii) defined and utilized safety metrics, (iv) common tools/simulators, and (v) stages of safety assessment addressed by each technique in the reviewed studies. As a result, we categorized safety assessment approaches into five groups: (1) model-based, (2) analytical-based, (3) data-driven, (4) experimental-based, and (5) hybrid approaches. We found that Monte Carlo simulation and Specific Operations Risk Assessment (SORA) are the most commonly used methods for safety assessment. We identified 42 metrics and classified them into four groups: (1) collision metrics, (2) performance metrics, (3) communication metrics, and (4) reliability metrics. Additionally, we identified 10 tools/simulators used for safety assessment. Finally, we observed that stage 5 (safety risk evaluation) of the safety assessment process is the most frequently covered in the studies reviewed.