Internet of things for automotive lets us enjoy a better in-car experience, and drives innovation of vehicular applications towards safer and more efficient transportation. This project focuses on solving the fundamental challenges of providing reliable and cost-effective fog computing services for the emerging vehicular applications that involve large-scale data-intensive crowdsensing and intelligent sensor fusion. Our proposal is vehicular fog computing, an innovative paradigm that brings computing power to the edge of the cellular network, and to predictably-routed commercial fleets like buses and taxis. The aim of this project is to develop a data-driven platform for effective capacity and resource management in vehicular fog computing. The expected outcomes will provide new insights on the design, implementation and deployment of vehicular fog computing.