Real-Time Path Recognition and Automatic Control of Educational Smart Cars

Autonomous driving represents a worldwide interest topic in the academic and industrial research shaping the future of the automotive development. In the last years, the largest suppliers of automotive embedded systems launched through their educational programs different contests that involve the development of autonomous small vehicles. This paper describes an algorithm for the real-time environmental perception for such an educational smart car. The data acquisition is performed using a camera sensor imposed by the specific contest. Based on the path recognition, a control algorithm for the smart vehicle is proposed. The algorithms are implemented in a microcontroller and the experimental results and learning outcomes are presented. The proposed algorithms can be applied by engineering students to better understand the self-driving perspective of the vehicles.