Assessment of Ant colony optimization on inducing non-uniform population distribution for UAS risk assessment in urban environments

Abstract

The proliferation of Unmanned Aircraft Systems (UAS) in commercial domains accentuates the imperative need for reliable safety measures, particularly within urban landscapes, during Unmanned Aerial Vehicle (UAV) operations. Contemporary frameworks, such as the Specific Operations Risk Assessment (SORA) proposed by the Joint Authorities for Rulemaking of Unmanned Systems (JARUS), predominantly use static methodologies for estimating population density, implicitly presuming uniform population distribution. Such assumptions potentially introduce inaccuracies in the quantification of third-party risks due to the dismissal of temporal fluctuations in population distribution. This study seeks to address this gap by employing Ant Colony Optimization (ACO), a bio-inspired algorithm traditionally used to solve optimization problems such as the Traveling Salesman Problem (TSP). Singapore serves as the case study in this context, where bus stops within neighbourhood subzones are assigned as network nodes. The ACO algorithm, coupled with an appropriate heuristic data input, is employed to induce non-uniformity in population distribution. Furthermore, the study introduces a metric, termed ’population exposure’, that acts as the heuristic data input to the ACO algorithm, and is devised to encapsulate dynamic patterns of population inflow and outflow. Preliminary findings, primarily conveyed through visualizations, show promising alignment with qualitative assumptions regarding population distribution about time and locale. This initial assessment could herald a nuanced approach to UAS operational risk assessments, accommodating the dynamic and non-uniform nature of urban population distribution.