College Students Build F1TENTH, the Fastest Self Driving Remote-Controlled Car
College students from the University of Waterloo have built F1TENTH, the fastest self driving remote-controlled car. This isn’t a project that anyone can build at home since the parts alone costs $5,000+, including anNVIDIA Jetson NX computing platform, Intel RealSense D3451i depth camera, VESC 6 MkV motor controller, and Hokuyo 10LX LiDAR module.
What makes this R/C car so fast? Well, it uses the best racing lines around a track by first mapping it with the LiDAR module as well as a software tool box. After mapping everything, the images were imported into Photoshop to create a more recognizable layout before a Euclidean Distance Transform finds the center line. An algorithm then uses that center line to determine the optimum racing lines for F1TENTH to take.
The raceline optimizer uses a different representation of the racetrack in order to find the optimal raceline: we need to represent the racetrack in terms of the centerline and track widths,” said the team.