Agent-based Modeling Approach for Operational Risk Assessment of Large-scale Low-altitude UAV Traffic

In this paper, we present an agent-based modeling framework for Unmanned Aerial Vehicle (UAV) operational risk assessment to support large-scale unmanned flights at low altitudes. The main objective of this risk model is to systematically evaluate real-time lethal risks posed to people on the ground by operating UAVs in urban low-altitude environments. Our framework comprises two key components: a ground risk model and an aerial risk model. The ground risk model assesses both the likelihood and severity of injuries to people on the ground caused by individual UAV failures, whereas the aerial risk model evaluates the probability of mid-air UAV collisions and their potential lethality. The proposed risk model is implemented in an agent-based simulation platform that generates UAV operations at multiple scales to evaluate their operational risks. Additionally, a re-pathfinding process is implemented for operations that exhibit unacceptable risk levels. The resulting outputs are analyzed and compared with a baseline scenario that does not incorporate risk assessment. The results reveal significant, previously unrecognized risks associated with large-scale UAV operations and underscore the necessity of developing a systematic risk assessment approach. Furthermore, the proposed model's limitations highlight key bottlenecks of UAV risk assessment, providing direction for future research.