Driving behavior in manned and unmanned learning vehicles and traffic ecosystems

Today's indispensable means of transport and transportation are important in human life. This is one of the important issues for drivers of safety and comfort during driving of vehicles. Drivers want their vehicles to be alerted and intervened in a possible malfunction. Otherwise, these faults can cause loss of life and property. Today, developed automobile systems are equipped with driver support systems. These driver support systems have been developed by analyzing past driving experiences and determining the deficiencies and incompleteness of the drivers. These systems are usually static and consistently perform the same interventions. For these systems, it is not a matter of who the driver is and what kind of mood he has. Actual behaviors and behaviors that can be achieved, and the answers and interventions they will give to these behaviors. In this thesis, firstly the data of a driver moving on a certain route is taken from the smartphone sensors using the developed android application, and a database of information about the driver is created. Later, this data is labeled, and it is determined by which labels the driver behaves in which region of the path. Following this step, the driver's overall driving profile is determined. When the driver of the general driving profile starts driving the car again, driving data is still being received and the results obtained are being improved. Thus, a system that is constantly evolving and updated is being created. When the driver changes, a system will be created that can create a driving profile within the new driver, and activate the advanced driver support system based on the profile of that driver. Thus, it is desired that the support systems of the driverless and non-driverless vehicles provide more effective support and have an effective intervention structure.