This study aims to investigate the factors and their complex interrelationships that influence key indicators of driving sustainability, with a focus on both safety and efficiency. A Structural Equation Model is developed using multi-source data to capture how unsustainable driving practices are shaped by the following three latent variables: driving volatility, route characteristics, and weather conditions. These latent variables are modeled using multiple observed indicators derived from smartphone sensors, OpenStreetMap, NASA satellite data, and weather sensors. The dataset consists of approximately 33,000 trips recorded via smartphone sensors in Athens, Greece, between March and May 2024. The model examines fuel consumption as a measure of efficiency, and four Surrogate Safety Measures (SSMs), speeding events, mobile phone use, and harsh acceleration and braking as safety-related variables. SEM enables estimation of both direct and indirect pathways, including how latent factors interact and how safer and more efficient driving behaviors may reinforce each other. This approach supports the generation of targeted insights to inform behavioral interventions for promoting safer and more efficient urban mobility.
| ID | pc657 |
| Presentation | |
| Full Text | |
| Tags |
