All technology is evolving towards automation. This also applies to vehicles that draw attention to themselves with their rapid development in the direction of autonomous driving. Vehicles with new assistance systems appear on the roads every day to help the driver while driving. This systems still require the driver to be present while driving and ready to grab the steering wheel. The question is, when will we witness the first vehicle that will be able to move on any road without a driver and even without a steering wheel and pedal.
This thesis presents the theoretical basis of autonomous vehicles, presents the levels of autonomous driving, describes sensors with which vehicles detect the environment, their advantages and disadvantages. As a practical example of the use of this technology, the thesis presents the design and construction of a self-driving car based on Raspberry Pi computer, that uses an optical sensor to detect the environment. We have optimized the vehicle, and for a successful ride, we raised the camera and set it to the optimal viewing angle. On the floor, however, we glued blue tape, which the camera detects as road lines.
We have found that the essential data for successful driving is the data detected by the camera, as the car decides where to turn based on this information. It is crucial that there are no objects of the same colour as the lines around the vehicle and that the camera captures enough data to have a wide enough angle of observation. Computer power, which processes all data, is also important.
|
