Nowadays the self-driving cars get more and more attention. The big tech companies, like Google, Tesla, invest a lot of money into researching this area.
Every year, BME is organizing a competition, where college students compete their skills and convert a remote control car into a self-driving autonomous car.
In a completed robot car, a number of sensors monitor the outside world, like distance measuring sensors, and a lot of systems need to work together, such as controllers, ABS, ASR, ACC, and respond very quickly to external changes. These parameters may vary during a fraction of a second, the amount of noise can be also significant, and it is almost impossible to make accurate measurements and to reach precise conclusions by empirical observation.
In my thesis, I'm introducing a system development that allows diagnostics of sensors and embedded software in a robot car. Making such a system in many ways is challenging, because it has to work in specific circumstances like low performance demand or mobility. These difficulties and the choices and tasks for solving them are presented in detail.