The 5th Annual F1/10th Competition
F1/10th provides a common hardware platform on which teams can rapidly develop and prototype the "brains" of an autonomous vehicle. The engineering challenges involve computer vision, optimal path planning and trajectory design, sensor fusion, and data acquisition, to name just a few. In short, a multi-disciplinary approach to engineering is needed.
Typically, the cost of entry to working on such an application (on a real-world scale) would prevent many research groups from tackling these important problems. The F1/10th platform (developed by Prof Rahul Mangharam and his team at UPenn) allows engineers to implement on the model cars with similar algorithms that would be used on real self-driving cars, but at a fraction of the cost and time.
The race took place in Carleton Commons on October 14th, with teams arriving the day before to test the track and calibrate their cars. Despite the jetlag, and powered by coffee and pizza, several teams stayed into the early hours of the morning. Race day consisted of three races. The first two races restricted the hardware requirements for all teams. Race one, won by the team from South Korea was a single-car-at-a-time fastest lap competition. Race two was a more technically demanding head-to-head race, won by the team from Poland, and the final race was the "open race", where the hardware restrictions were lifted, and the winners were Team Dzik from the Polish Academy of Sciences.
Next year, Columbia plans to enter its first ever team into the race, led by incoming EE Professor James Anderson, which takes place at the 21st IFAC World Congress in Berlin.
Students (undegrad and grad students) that are interested in learning more about the Columbia team can contact Prof. Anderson here.