Implementing and Tuning an Autonomous Racing Car Testbed
Ref: CISTER-TR-191209 Publication Date: 26, Nov, 2019
Implementing and Tuning an Autonomous Racing Car TestbedRef: CISTER-TR-191209 Publication Date: 26, Nov, 2019
Achieving safe autonomous driving is far from a vision at present days, with many examples like Uber, Google and the most famous of all Tesla, as they successfully deployed self driving cars around the world. Researchers and engineers have been putting tremendous efforts and will continue to do so in the following years into developing safe and precise control algorithms and technologies that will be included in future self driving cars. Besides these well known autonomous car deployments, some focus has also been put into autonomous racing competitions, for example the Roborace. The fact is that although significant progress that has been made, testing on real size cars in real environments requires immense financial support, making it impossible for many research groups to enter the game. Consequently, interesting alternatives appeared, such as the F1 Tenth, which challenges students, researchers and engineers to embrace in a low cost autonomous racing competition while developing control algorithms, that rely on sensors and strategies used in real life applications. This thesis focus on the comparison of different control algorithms and their effectiveness, that are present in a racing aspect of the F1 Tenth competition. In this thesis, efforts were put into developing a robotic autonomous car, relying on Robot Operative System, ROS, that not only meet the specifications from the F1 Tenth rules, but also allowed to establish a testbed for different future autonomous driving research.
Master Thesis, ISEP.
Notes: Orientação científica: Ricardo Severino