AutoRally AN OPEN PLATFORM FOR AGGRESSIVE AUTONOMOUS DRIVING

被引:62
作者
Goldfain, Brian [1 ,2 ]
Drews, Paul [3 ]
You, Changxi [4 ]
Barulic, Matthew [5 ]
Velev, Orlin [6 ]
Tsiotras, Panagiotis [7 ,8 ]
Rehg, James M. [1 ,9 ,10 ,11 ]
机构
[1] Georgia Inst Technol, Sch Interact Comp, Atlanta, GA 30332 USA
[2] Georgia Inst Technol, 85 5th St NW,Off 218a, Atlanta, GA 30308 USA
[3] Georgia Inst Technol, Sch Elect & Comp Engn, Atlanta, GA 30332 USA
[4] Georgia Inst Technol, Sch Aerosp Engn, Atlanta, GA 30332 USA
[5] Wheego Technol Inc, Atlanta, GA USA
[6] Space Explorat Technol SpaceX, Hawthorne, CA USA
[7] Georgia Inst Technol, Inst Robot & Intelligent Machines, Atlanta, GA 30332 USA
[8] Georgia Inst Technol, Dynam & Control Syst Lab, Atlanta, GA 30332 USA
[9] Georgia Inst Technol, Ctr Behav Imaging, Atlanta, GA 30332 USA
[10] Georgia Inst Technol, Ctr Computat Hlth, Atlanta, GA 30332 USA
[11] Georgia Inst Technol, Computat Percept Lab, Atlanta, GA 30332 USA
来源
IEEE CONTROL SYSTEMS MAGAZINE | 2019年 / 39卷 / 01期
关键词
45;
D O I
10.1109/MCS.2018.2876958
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
AutoRally is an open-source, 1:5-scale autonomous vehicle testbed for students, researchers, and engineers who are interested in autonomous vehicle technologies. It is designed with robustness and ease of use in mind. At 1 m in length, weighing 22 kg, and with a top speed of 90 km/h, the platform is large enough to host powerful on-board computing and sensing and run state-of-the-art algorithms. At the same time, it is simple and small enough to be maintained and operated by two people, all while providing the capability of exploring driving scenarios including drifting, jumping, high-speed driving, and multivehicle interactions. Build instructions are publicly available, along with a parts list, computer-aided design models for fabricating custom components, and operating procedures. The platform uses the Robot Operating System (ROS) and can be programmed in Python or C++. Tutorials, reference algorithms, a Gazebo-based simulation environment, and a data set structured as ROS bag files are available from the AutoRally website. The fleet of six AutoRally platforms at Georgia Tech have been used to demonstrate control, perception, and estimation research in a high-speed, off-road driving domain. To date, the fleet has driven hundreds of kilometers autonomously at the Georgia Tech Autonomous Racing Facility.
引用
收藏
页码:26 / 55
页数:30
相关论文
共 42 条
[1]  
[Anonymous], AUT VEH SELF DRIV VE
[2]  
[Anonymous], 2017, C ROB LEARN
[3]  
[Anonymous], 2014, TAX DEF TERMS REL ON
[4]   SegNet: A Deep Convolutional Encoder-Decoder Architecture for Image Segmentation [J].
Badrinarayanan, Vijay ;
Kendall, Alex ;
Cipolla, Roberto .
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2017, 39 (12) :2481-2495
[5]  
Bertoncello M., 2015, Ten ways autonomous driving could redefine the automotive world
[6]  
Burhaumudin M. S., 2012, INT J RES ENG TECHNO, V1, P139
[7]   DeepLab: Semantic Image Segmentation with Deep Convolutional Nets, Atrous Convolution, and Fully Connected CRFs [J].
Chen, Liang-Chieh ;
Papandreou, George ;
Kokkinos, Iasonas ;
Murphy, Kevin ;
Yuille, Alan L. .
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 2018, 40 (04) :834-848
[8]  
Cutler M, 2014, IEEE INT CONF ROBOT, P3888, DOI 10.1109/ICRA.2014.6907423
[9]  
Dellaert F., 2012, Factor graphs and GTSAM: A hands-on introduction
[10]   Square root SAM: Simultaneous localization and mapping via square root information smoothing [J].
Dellaert, Frank ;
Kaess, Michael .
INTERNATIONAL JOURNAL OF ROBOTICS RESEARCH, 2006, 25 (12) :1181-1203