| Demand of automobiles are significantly growing despite various factors, thereby, steadily increasing the average number of vehicles on the road. Increase in the number of vehicles, subsequently increases the risk of collisions, inturn characterized by the driving behavior. Driving behavior is influenced by factors like class of vehicle, road conditions and vehicle maneuvering by the driver. Rapidly growing mobile technology and use of smartphones, embedded with in-built sensors, provide great scope of constant development of assistance systems considering the safety of the driver, which is possible by integrating the information obtained from the vehicle on-board sensors. Our research aims at learning a complete vehicle system comprising of vehicle, human and road, by employing driving patterns obtained from the sensor data in order to develop better systems of safety and alerts altogether.;The thesis focusses on utilizing all the various data recorded by the embedded sensors built in a smartphone to understand the vehicle motion and dynamics. This is followed by studying various impacts of collision events, types and signatures, which can potentially be integrated in a prototype framework to detect variations in order to alert drivers and emergency responders in an event of collision. |
