Detecção de padrões anômalos na mobilidade urbana por meio de indexação espacial com células H3

Abstract

This research investigates the application of the H3 Hexagonal Indexing System in the analysis of urban mobility patterns and fraud detection in taxi services. Using data from yellow taxi trips in New York City (January 2024), the study employs a deductive, quantitative, and applied methodology, implementing spatial analysis techniques with hierarchical hexagonal cells to identify regular behaviors and anomalies. The results demonstrate the superiority of H3 compared to traditional methods, presenting 37% more precision in fraud detection and enabling the identification of specific clusters with irregular practices, such as short trips with high fares and records of impossible speeds. The analysis revealed significant geographic concentrations of anomalies in areas such as airports and tourist zones, as well as recurring temporal patterns, evidencing deliberate and systematic behaviors. It is concluded that the proposed methodology significantly contributes to the supervision and improvement of urban transport services, with potential application in other metropolitan contexts.