A Simulation Benchmark for Autonomous Racing with Large-Scale Human Data

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Remonda, Adrian, Hansen, Nicklas, Raji, Ayoub, Musiu, Nicola, Bertogna, Marko, Veas, Eduardo, Wang, Xiaolong
Formato: Preprint
Publicado: 2024
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866912738492547072
author Remonda, Adrian
Hansen, Nicklas
Raji, Ayoub
Musiu, Nicola
Bertogna, Marko
Veas, Eduardo
Wang, Xiaolong
author_facet Remonda, Adrian
Hansen, Nicklas
Raji, Ayoub
Musiu, Nicola
Bertogna, Marko
Veas, Eduardo
Wang, Xiaolong
contents Despite the availability of international prize-money competitions, scaled vehicles, and simulation environments, research on autonomous racing and the control of sports cars operating close to the limit of handling has been limited by the high costs of vehicle acquisition and management, as well as the limited physics accuracy of open-source simulators. In this paper, we propose a racing simulation platform based on the simulator Assetto Corsa to test, validate, and benchmark autonomous driving algorithms, including reinforcement learning (RL) and classical Model Predictive Control (MPC), in realistic and challenging scenarios. Our contributions include the development of this simulation platform, several state-of-the-art algorithms tailored to the racing environment, and a comprehensive dataset collected from human drivers. Additionally, we evaluate algorithms in the offline RL setting. All the necessary code (including environment and benchmarks), working examples, datasets, and videos are publicly released and can be found at: https://assetto-corsa-gym.github.io
format Preprint
id arxiv_https___arxiv_org_abs_2407_16680
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle A Simulation Benchmark for Autonomous Racing with Large-Scale Human Data
Remonda, Adrian
Hansen, Nicklas
Raji, Ayoub
Musiu, Nicola
Bertogna, Marko
Veas, Eduardo
Wang, Xiaolong
Robotics
Machine Learning
Despite the availability of international prize-money competitions, scaled vehicles, and simulation environments, research on autonomous racing and the control of sports cars operating close to the limit of handling has been limited by the high costs of vehicle acquisition and management, as well as the limited physics accuracy of open-source simulators. In this paper, we propose a racing simulation platform based on the simulator Assetto Corsa to test, validate, and benchmark autonomous driving algorithms, including reinforcement learning (RL) and classical Model Predictive Control (MPC), in realistic and challenging scenarios. Our contributions include the development of this simulation platform, several state-of-the-art algorithms tailored to the racing environment, and a comprehensive dataset collected from human drivers. Additionally, we evaluate algorithms in the offline RL setting. All the necessary code (including environment and benchmarks), working examples, datasets, and videos are publicly released and can be found at: https://assetto-corsa-gym.github.io
title A Simulation Benchmark for Autonomous Racing with Large-Scale Human Data
topic Robotics
Machine Learning
url https://arxiv.org/abs/2407.16680