Arun Das

Arun NASA

Research:

PhD: I am investigating ways to apply information theory to the SLAM problem. Specifically, I am looking at how to quantify the information provided by observed measurements, and how the measurement information relates to the observability and sensitivity of the vehicle state, especially when the system is augmented with additional inertial and GNSS sensors. Quantification of the measurement information is a powerful ability, as it allows us to predict which features of the environment will provide the most useful data for estimating the vehicle motion. Furthermore, this research develops a generalized information theoretic framework which will be used to analyse and improve existing SLAM algorithms, and also provide complementary approaches for other important mapping functions, such as key-frame selection and sensor reconfiguration.

Masters: My masters work looked at pointcloud based SLAM and scan registration, where I specifically looked at methods to overcome the shortcomings of the NDT algorithm, in both 2D and 3D scenarios through the Multi-Scale k-Means NDT (MSKM-NDT) and the Segmented Greedy Cluster NDT (SGC-NDT) algorithms. These methods significantly reduce the number of Gaussian clusters required for scan registration, and propose a smooth and continuous cost function that guarantees convergence of the optimization.  Master’s Thesis.

 

Projects:

NASA Sample Return Robot Challenge (NSRRC):

The NSRRC is a NASA Centennial Challenge which offers a $1.5 Million dollar prize to teams which are able to demonstrate autonomous navigation and sample collection and retrieval over a varied 80,000 square meter sized terrain.  As part of this project, I led the team which developed a real-time, 3D laser-based, SLAM algorithm. We successfully developed a strategy capable of localizing the robot to within 10 cm accuracy in a large, forested area. A video of the 2D obstacle map can be seen below:

  An overview of the 2014 robot:

Selected Publications:

Journal Publications:

  • A. Das, S. Waslander, and M. Haley. 3D Mapping in Tunnel Environments: Maximizing Mission Range. Progress and Communication in Sciences, 1(2):46–51, 2014.
  • A. Das and S. Waslander. Scan registration using segmented region growing NDT. International Journal of Robotics Research, 33(13):1645-1663, 2014.
  • A. Das, M. Diu, N. Mathew, C. Scharfenberger, J. Servos, A. Wong, J. Zelek, D. Clausi, and S. Waslander. Mapping, planning and object detection strategies for autonomous exploration. Journal of Field Robotics: Special Issue on Space Robotics II, 31(1):75–106, 2013.

Conference Publications:

  • A. Das and S. Waslander. Entropy Based Keyframe Selection for Multi-Camera Visual SLAM. In IEEE International Conference on Intelligent Robots and Systems (IROS), Hamburg, Germany, Oct 2015.
  • A. Das, D. Kumar, A. El Bably, S.L. Waslander. Taming the North: Multi-Camera Parallel Tracking and Mapping in Snow-Laden Environments. In Field and Service Robotics, Toronto, ON, Canada, 2015.
  • K. Ling, D. Chow, A. Das, and S.L. Waslander. Autonomous maritime landings for low-cost VTOL aerial vehicles. In Canadian Conference on Computer and Robot Vision (CRV), pages 32–39, Washington, DC, USA, 2014.
  • A. Das and S. Waslander. Outlier rejection for visual odometry using parity space methods. In IEEE International Conference on Robotics and Automation (ICRA), pages 3613–3618, Hong Kong, China, 2014.
  • A. Das, J. Servos, and S. Waslander. 3D scan registration using the normal distributions transform with ground segmentation and point cloud clustering. In IEEE International Conference on Robotics and Automation (ICRA), pages 2199–2204, Karlsruhe, Germany, May 2013.
  • A. Das and S. Waslander. Scan registration with multi-scale k-means normal distributions transform. In IEEE International Conference on Intelligent Robots and Systems (IROS), pages 2705–2710, Vilamoura, Algarve, Portugal, Oct 2012.

Selected Awards:

  • Alexander Graham Bell Canadian Graduate Scholarship – Doctoral, Natural Sciences and Engineering Research Council of Canada, Waterloo, ON. (2015-2017)
  • President’s Graduate Scholarship, University of Waterloo, Waterloo, ON. (2015-2017)
  • Carl A. Pollock Postgraduate Fellowship, University of Waterloo, Waterloo, ON. (2014-2015)
  • President’s Graduate Scholarship, University of Waterloo, Waterloo, ON. (2013-2014).
  • Ontario Graduate Scholarship, Ontario Ministry of Training, Colleges and Universities, Waterloo, ON. (2013-2014)
  • NASA Sample Return Robot Challenge People’s Choice Award, NASA Centennial Challenges, Worcester, MA. (2013)
  • BlackBerry Developer Conference Lightning Pitch Winner, BlackBerry Developers’ Conference, Waterloo, ON. (2012)
  • President’s Graduate Scholarship, University of Waterloo, Waterloo, ON. (2011-2012)
  • Ontario Graduate Scholarship (Declined), Ontario Ministry of Training, Colleges and Universities, Waterloo, ON. (2011-2012)
  • Alexander Graham Bell Canadian Graduate Scholarship – Masters, Natural Sciences and Engineering Research Council of Canada, Waterloo, ON. (2011-2012)
  • President’s Graduate Scholarship, University of Waterloo, Waterloo, ON. (2010-2011)
  • Ontario Graduate Scholarship, Ontario Ministry of Training, Colleges and Universities, Waterloo, ON. (2010-2011)

Contact:

Address:

Arun Das
Department of Mechanical and Mechatronics Engineeering
University of Waterloo
Waterloo, ON, Canada
N2L 3G1

e-mail:

arun[dot]das[at]uwaterloo[dot]ca

office:

E5-3018

phone: 

(519)888-4567×31404

Search

Waterloo Autonomous Vehicles Lab  Contact Us