Assessing ant colony optimization for inducing non-uniform population distribution for UAS risk assessment in urban environments

• Main Author: Gracious, John
• Other Authors: Low Kin Huat
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2023
• Subjects: Engineering::Aeronautical engineering::Accidents and air safety
• Online Access: https://hdl.handle.net/10356/167971
• Institution: Nanyang Technological University

Unmanned Aircraft Systems (UAS), commonly referred to as drones, were initially developed for military purposes to conduct reconnaissance missions without risking the lives of pilots. Advancements in commercial-grade technology have made it possible to re-scope the purpose of drones for commercial operations such as package delivery and aerial photography. However, safety of UAS operations continues to pose a significant obstacle, particularly in urbanized environments. The Joint Authorities for Rulemaking of Unmanned Systems (JARUS) recommends the Specific Operations Risk Assessment (SORA) framework for quantifying third-party risks associated with UAS operations. Nevertheless, ground risk assessments remain elusive due to dynamic factors such as non-uniform population distribution. Contemporary approaches for assessing ground risk involve the use of a static measure for estimating population density. This method assumes that the total population within a city or subzone is uniformly distributed. Moreover, these approaches neglect to account for the dynamic shifts in population distribution over time, resulting in an imprecise assessment of third-party risks. Hence, the main objective of this research is to examine the use of Ant Colony Optimization (ACO) to induce non-uniform population distribution in urbanized environments. The novelty of this study lies in its distinctive approach for quantifying non-uniform population distribution through the utilization of ACO, a technique conventionally used for solving the Traveling Salesman Problem (TSP). Specifically, this investigation utilizes public bus stops as nodes and applies the ACO algorithm to analyse the bus transportation network within subzones in Singapore. In addition, an original metric called ‘population exposure’ is introduced to capture people's inflow and outflow patterns from different areas through these public bus stops. This serves as heuristic information during network graph generation and enables non-uniformity while facilitating temporal population analysis. In essence, this allows for preliminary assessments on temporal fluctuations in population movement. The results of this study are mainly visualizations. The results are promising as it conforms to qualitative assumptions typically made about certain subzone at certain times of day. Overall, the study is preliminary, but further research in this area may facilitate a refined analysis of population movement in UAS risk assessments.