Public road vehicles can be considered to be totally autonomous if no human interference is needed for their operation. Modern autonomous cars, relying on data provided by certain sensors, such as radars and video cameras, are able to detect and categorize objects in their environment. Besides this, their aims are multiple: to stay on the optimal track, to obey traffic rules, to boost the driving experience and, most importantly, to serve the driver’s and the passengers’ safety.
How can road safety be increased? How can autonomous vehicles remain compatible with the other, non-autonomous participants of traffic?
As the prevention of accidents is of key importance, the development of autonomous vehicles is highly motivated. Driver Assistance (DA) forms the basis of such engineering endeavours.
Concerning the validation of autonomous vehicles, the tests of the New Car Assessment Programme (NCAP) are the most renowned. Can the car stop before its target? If not, what speed reduction can be achieved by the DA system?
My thesis focuses on developing a verification software to serve the goals mentioned above. My program collects the necessary input data from the resimulated measurement files, such as the speed of the tested vehicle, the braking force required by the DA system, the distance from the target vehicles and its speed, and finally calculates the braking force of the vehicle. The software specifies whether the vehicle can stop in time or hits the target.
The collected information could be of great help for vehicle engineers because they could find the optimal settings during software development, by the help of which the vehicles could reach the highest possible scores in resimulated NCAP tests.