AutoRally An open platform for aggressive autonomous driving

TL;DR

AutoRally is an accessible, scalable robotics platform designed for autonomous vehicle research, enabling easy construction, operation, and data collection for diverse autonomous driving studies.

Contribution

It introduces a comprehensive, open-source platform with detailed documentation, algorithms, and datasets, facilitating research in aggressive autonomous driving with minimal technical barriers.

Findings

Successful real-world data collection with six robots

Effective offline and online state estimation algorithms

Demonstrated robustness and versatility of the platform

Abstract

This article presents AutoRally, a 1$:$5 scale robotics testbed for autonomous vehicle research. AutoRally is designed for robustness, ease of use, and reproducibility, so that a team of two people with limited knowledge of mechanical engineering, electrical engineering, and computer science can construct and then operate the testbed to collect real world autonomous driving data in whatever domain they wish to study. Complete documentation to construct and operate the platform is available online along with tutorials, example controllers, and a driving dataset collected at the Georgia Tech Autonomous Racing Facility. Offline estimation algorithms are used to determine parameters for physics-based dynamics models using an adaptive limited memory joint state unscented Kalman filter. Online vehicle state estimation using a factor graph optimization scheme and a convolutional neural network for semantic segmentation of drivable surface are presented. All algorithms are tested with real world data from the fleet of six AutoRally robots at the Georgia Tech Autonomous Racing Facility tracks, and serve as a demonstration of the robot$'$s capabilities.