Autonomous Racing: A Comparison of SLAM Algorithms for Large Scale Outdoor Environments

The task of simultaneous localization and mapping (SLAM) is a widely studied field in robotics research in the last decades. The goal of SLAM is to create an accurate map of the environment considering uncertainties in the pose as well as the environmental perception of the robot. Historically SLAM algorithms are applied in the field of indoor robotics. Recent developments in the area of autonomous driving surge a focus for SLAM applications in large scale outdoor environments. Two notable open source SLAM software packages are Gmapping and Google Cartographer. This paper focuses on a qualitative comparison of the aforementioned algorithms for such a scenario. We discuss the underlying algorithmic differences of the two packages. This serves as the foundation to present the SLAM results for different parameter configurations. We evaluate the accuracy of the resulting maps and the respective computational limitations. The maps are further evaluated against manually measured ground truth track boundaries. We show that the existing approaches can be adapted to large-scale outdoor environments.

[1]  Wolfram Burgard,et al.  3-D Mapping With an RGB-D Camera , 2014, IEEE Transactions on Robotics.

[2]  David Portugal,et al.  An evaluation of 2D SLAM techniques available in Robot Operating System , 2013, 2013 IEEE International Symposium on Safety, Security, and Rescue Robotics (SSRR).

[3]  Sebastien Glaser,et al.  Simultaneous Localization and Mapping: A Survey of Current Trends in Autonomous Driving , 2017, IEEE Transactions on Intelligent Vehicles.

[4]  Markus Lienkamp,et al.  What can we learn from autonomous level-5 motorsport? , 2018, Proceedings.

[5]  Wolfgang Hess,et al.  Real-time loop closure in 2D LIDAR SLAM , 2016, 2016 IEEE International Conference on Robotics and Automation (ICRA).

[6]  Edwin Olson,et al.  Real-time correlative scan matching , 2009, 2009 IEEE International Conference on Robotics and Automation.

[7]  Michael Bosse,et al.  Map Matching and Data Association for Large-Scale Two-dimensional Laser Scan-based SLAM , 2008, Int. J. Robotics Res..

[8]  Stefan Kohlbrecher,et al.  A flexible and scalable SLAM system with full 3D motion estimation , 2011, 2011 IEEE International Symposium on Safety, Security, and Rescue Robotics.

[9]  Wolfram Burgard,et al.  Improved Techniques for Grid Mapping With Rao-Blackwellized Particle Filters , 2007, IEEE Transactions on Robotics.

[10]  Andreas Nüchter,et al.  IMPROVING GOOGLE'S CARTOGRAPHER 3D MAPPING BY CONTINUOUS-TIME SLAM , 2017 .

[11]  Sebastian Thrun,et al.  FastSLAM 2.0: An Improved Particle Filtering Algorithm for Simultaneous Localization and Mapping that Provably Converges , 2003, IJCAI.

[12]  Sebastian Thrun,et al.  Probabilistic robotics , 2002, CACM.

[13]  Patrick Sebastian,et al.  A quantitative study of tuning ROS gmapping parameters and their effect on performing indoor 2D SLAM , 2016, 2016 2nd IEEE International Symposium on Robotics and Manufacturing Automation (ROMA).

[14]  Hans P. Moravec.  Sensor Fusion in Certainty Grids for Mobile Robots , 1988, AI Mag..

[15]  Sebastian Thrun,et al.  Perspectives on standardization in mobile robot programming: the Carnegie Mellon Navigation (CARMEN) Toolkit , 2003, Proceedings 2003 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2003) (Cat. No.03CH37453).

[16]  Juan D. Tardós,et al.  ORB-SLAM2: An Open-Source SLAM System for Monocular, Stereo, and RGB-D Cameras , 2016, IEEE Transactions on Robotics.

[17]  Marcelo H. Ang,et al.  Perception, Planning, Control, and Coordination for Autonomous Vehicles , 2017 .

[18]  S. M. Mamduh,et al.  Performance Analysis of the Microsoft Kinect Sensor for 2D Simultaneous Localization and Mapping (SLAM) Techniques , 2014, Sensors.

[19]  F Gustafsson,et al.  Particle filter theory and practice with positioning applications , 2010, IEEE Aerospace and Electronic Systems Magazine.