This repository contains maps from over 20 real race tracks (mainly F1 and DTM) downscaled on 1:10 for the usage in the F1TENTH Gym and F1TENTH Simulator.
The collected research papers in this repository were used to create the paper "Autonomous Vehicles on the Edge: A Survey on Autonomous Vehicle Racing". If you find the information in this repository useful we would be happy if you cite it based on the following definition:
@article{Betz2022_RacingSurvey,
title = "Autonomous vehicles on the edge: A survey on autonomous vehicle
racing",
author = "Betz, Johannes and Zheng, Hongrui and Liniger, Alexander and
Rosolia, Ugo and Karle, Phillip and Behl, Madhur and Krovi,
Venkat and Mangharam, Rahul",
journal = "IEEE Open J. Intell. Transp. Syst.",
publisher = "Institute of Electrical and Electronics Engineers (IEEE)",
volume = 3,
pages = "458--488",
year = 2022,
doi = "10.1109/ojits.2022.3181510"
}
- Austin, USA (F1)
- Brands Hatch, UK (DTM)
- Budapest, Hungary (F1)
- Catalunya, Spain (F1)
- Hockenheim, Germany (F1, DTM)
- IMS Indianapolis Motor Speedway, USA (IndyCar)
- Melbourne, Australia (F1)
- Mexico City, Mexico (F1)
- Montreal, Canada (F1)
- Monza, Italy (F1)
- Moscow Raceway, Russia (DTM)
- Nuerburgring, Germany (DTM)
- Oschersleben, Germany (DTM)
- Sakhir, Bahrain (F1)
- Sao Paulo, Brazil (F1)
- Sepang, Malaysia (F1)
- Shanghai, China (F1)
- Silverstone, UK (F1)
- Sochi, Russia (F1)
- Spa, Belgium (F1)
- Spielberg, Austria (F1)
- Yas Marina, Abu Dhabi (F1)
- Zandvoort, Netherlands (DTM)
This repository is based on the f1 racketrack database which includes the racetrack information (inner bounds, outer bounds, track width) of 20 real racetracks all around the world. The original center lines on the real racetracks were fetched as GPS points from the OpenStreetMap project (https://www.openstreetmap.org). We applied a smoothing algorithm to the center lines. The track widths were extracted from satellite images using an image processing algorithm. Afterwards the data was downscaled 1:10 with a fixed track width of 2,20m so it can be used in the F1TENTH environments.
- maps: The tracks are displayed as occupancy grid maps based on a .png file. That maps have a corresponding .yaml file that includes both resolution and origin of the maps. These maps can be used in both the F1TENTH Gym and F1TENTH Simulator.
- centerline: [x_m, y_m, w_tr_right_m, w_tr_left_m] These files contain the center line (x, y) and the track widths to the right (w_tr_right) and left (w_tr_left). The center lines were smoothed. Therefore, they do not lie perfectly in the middle of the track anymore.
- raceline: [s_m; x_m; y_m; psi_rad; kappa_radpm; vx_mps; ax_mps2] The provided race lines were calculated minimizing the summed curvature. They lie within the track boundaries. The raceline files contain the s-value along the raceline, the racline positions (x,y,psi) the optimal curvature (kappa), the optimal longitudinal veloctiy (vx) and longitudinal acceleration (ax) for the specific raceline position.
- waypoints_DonkeySim: This file contains the optimal raceline provided for the usage in the Donkey Simulator.
The heading in the raceline files is using a coordinate system from 0-2pi. This is mainly for the usage in the F1TENTH gym. If you want to use another coordinate system conversion we also provide a conversion script that transforms the racelines created by optimization to the into the coordinate system used by your simulation.
The conversion script is convert.py, and takes one argument --pattern for the pattern used by glob to find all files matching the pattern. The default pattern is */*raceline.csv/, which correspond to all .csv files that was appended by raceline in the file name.
The original f1 racetrack database was created by Alexander Heilmeier from the Institute of Automotive Technology at the Technical University of Munich.