Investigation and Implementation of Self-driving Functions on a Model Car

OData support
Supervisor:
Kiss Domokos
Department of Automation and Applied Informatics

Considering the traffic today, more than a million people die each year and additionally 20 to 50 million people have been injured in road accidents. This number is equal to more than ten aircraft accidents every day. Of course, the two traffic types differ from each other in many aspects, such as traffic density, central control, or disturbing objects, but the goal is the same, to reduce the number of fatal accidents. The main cause of the accidents - in most cases - can be related to human failure. In the future in order to make road transport safer, the systems of the vehicle must give help for the drivers during their journey, therefore today's even more popular driving assistance systems and later self-driving systems need to be developed.

In my thesis I have worked on the implementation of several self-driving functions. In order to demonstrate the tasks in real life, I handled them as a mobile robotic problem, just as in case of industrial development of autonomous vehicles.

In the thesis I will introduce current developments and the classification of self-driving cars. I'm providing an insight into the world of mobile robotic problems and approaches. I describe the robot platform that I'm using and present the sensing, information extraction and motion control, as well as the methods applied for localization and mapping tasks. As a result, the car can demonstrate several self-driving features, namely lane keeping and intersection recognition, as well as localization and mapping abilities.

Downloads

Please sign in to download the files of this thesis.