Bringing Big Data into the Car: Does it Scale?

摘要

The increasing velocity of big data captured by various sensors and processed in real-time offers support for a range of new application domains. For car information systems (CIS), data from different sources including IoT needs to be combined to offer an adequate service to the user. In this paper, we introduce a novel CIS big data-centric architecture based on a smart streaming infrastructure integrating data source in and outside of the car. We have created a prototype implementation of this architecture and run several experiments to validate the quality of our solution. Especially, we have examined the fault tolerance of the architecture by systematically introducing failures and evaluating their effects on the car information system. The experimental results show that our solution for a smart data based car information system is both scalable and fault tolerant.