Quando chega? análise de previsibilidade de tempos de espera em transporte público urbano utilizando modelos mistos

Abstract

This study presents an exploratory analysis of open urban public transportation data and proposes models for predicting passenger waiting times at bus stops. To achieve this, mixed models and confidence intervals were employed to identify operational patterns in a case study focused on the city of Curitiba. The results show that it is possible to develop predictive models incrementally, taking into account the specific characteristics of the analyzed bus lines and adapting the models to the needs and realities of the available data. The contributions of this study include: (1) the implementation of linear mixed models and their potential benefits for both users and system operators; (2) the integration of methods and tools specifically designed to handle public transportation data; (3) a discussion on how model outputs can be used to assess service quality; and (4) recommendations on relevant variables to be included in the models and their impact on waiting times.