Obstacle detection, classification and avoidance for F1tenth autonomous racing

Co-supervisor: Japie Engelbrecht

F1tenth is an international community performing autonomous systems research through building and racing of 1:10 scale F1 vehicles. The common goal is to race around tracks as fast as possible – usually first in simulation and then in the real world. This provides a platform for addressing many research and systems engineering questions and solving both theoretical and practical problems. The current approaches to autonomous racing ranges from more conventional control approaches (e.g., model predictive control) to data-driven solutions involving machine learning (e.g., reinforcement learning and imitation learning).

In the case of multi-vehicle racing, a central challenge is avoiding collisions with static and dynamic obstacles, in the form of track boundaries and other vehicles, respectively. It is therefore very important to accurately detect and classify these in most driving paradigms, as opposed to making decisions based directly on raw sensor data. The focus of this project will be on the perception component of the overall architecture, which takes a global trajectory as input and needs to output suitable information to the downstream (control) systems that in turn need to adapt accordingly. The experiments will be performed in simulation, but then transferred onto physical vehicles using ROS.