Estimation of a Mobility Matrix for the City of Cuenca Based on Human Flow Data

This study presents a novel methodology to estimate an origin-destination (OD) mobility matrix for the city of Cuenca, Ecuador, using geospatial data obtained via the GPS Logger mobile application and machine learning techniques. From a sample of 380 individuals, urban travel patterns were identified through the analysis of trip micro-cycles. Travel routes were classified using a decision tree model, achieving an accuracy of 95.7%, and indicating that the most frequented zones correspond to educational and commercial areas. The findings reveal that only 22% of trips involve public transportation, while 78% rely on private vehicles, emphasizing the urgent need to promote sustainable transport systems. The proposed methodology demonstrates significant potential for urban planning and data-informed policies.