Research On Context Model And Middleware In Smart Car

Cars are become more and more frequently used places in our daily life. Yet driving also brings problems, such as congestion and accident. Drivers want their driving to be easier with less workload and less chance of getting injured. Now information and communication technologies make it no longer a dream to build smart car. Advanced Driver Assistant System (ADAS) have made a rapid progress over the past two decades. Pervasive computing provides theoretical and technical support for smart car to improve transportation safety, efficiency, and comfort.Context-awareness is the key feature of smart cars. In-vehicle context aware systems aim at taking into account more contextual information related to the driving task in order to produce adapted or customized action. Highly integrated contexts will enable smart car to provide appropriate service and correct suggestion for the driver, such as assisting the driver to prevent accidents, providing drivers with real-time information about the road network to avoid congestion, and optimizing a journey to reduce overall efficiency.In this thesis, the contributions are four folds:1) A novel context model is proposed, including sensor layer, context atom layer and context situation layer, which can represent complex driving environment;2) Context reasoning mechanism is developed according to the context model, responsible for fusion of context situations from context atoms;3) A middleware of the smart car is implemented, including low level context acquisition and sharing, high level situation recognition and prediction, and finally services for enhancement of the safety for all the road users;4) A smart car prototype is built to provide runtime environment for different services and to implement the capability of context-awareness.