John J. Leonard
Professor of Mechanical and Ocean Engineering
Teaching
In Fall 2008, I will teach 2.166 Probabilistic Techniques for Mobile Robotics (to be held MW2.30-4 in room 1-371). Here's the description from the MIT Course Catalog:Theory and application of probabilistic techniques for autonomous mobile robotics. Topics include probabilistic state estimation and decision making for mobile robots; stochastic representations of the environment; dynamic models and sensor models for mobile robots; algorithms for mapping and localization; planning and control in the presence of uncertainty; cooperative operation of multiple mobile robots; mobile sensor networks; application to autonomous marine (underwater and floating), ground, and air vehicles.
Research
My recent research has addressed the problem of simultaneous localization and mapping (SLAM) autonomous mobile robots. The problem of SLAM is stated as follows: starting from an initial position, a mobile robot travels through a sequence of positions and obtains a set of sensor measurements at each position. The goal is for the mobile robot to process the sensor data to produce an estimate of its position while concurrently building a map of the environment. While the problem of SLAM is deceptively easy to state, it presents many theoretical challenges. The problem is also of great practical importance; if a robust, general-purpose solution to SLAM can be found, then many new applications of mobile robotics will become possible.
Under funding from the Sea Grant College Program and the Office of Naval Research, my research group is developing new SLAM algorithms for AUVs using sonar. In 2002, 2004, and 2005, we participated in a series of AUV experiments in Italy performed in collaboration with NATO SACLANT Undersea Research Centre. A video from the GOATS-2002 experiment is available here. Our work makes use of two new Odyssey III AUVs from Bluefin Robotics. (This work is being performed in collaboration with Prof. Henrik Schmidt of the MIT acoustics group and Prof. Chrys Chryssostomidis of the MIT Sea Grant Collge Program.)
Our SLAM research has applicability a wide range of robots operating in a diverse set of environment, making use of laser, sonar, and/or visual sensing. One of our goals has been to enable a robot to autonomously navigate a large-scale environment, such as the buildings of the MIT campus. A video of mapping results for the MIT campus is available at here. (This work is joint with Paul Newman, Mike Bosse, Seth Teller, Juan Domingo Tardós, and José Neira.)
The primary goal of our ongoing research is to pursue the challenge of persistent autonomy --- the capability for one or more robots to operate robustly for days, weeks and months at a time with minimal human supervision, in complex, dynamic environments. Taking the limit as time goes to infinity poses difficult challenges to our algorithms, but this is imperative for many applications of autonomous mobile robots. For example, security missions require the capability for robots to build and maintain maps of large areas, detecting changes and correcting their internal representations to maintain currency with the world. These capabilities are beyond the match of today's robots.
We believe that the critical challenges for future research in this area are two-fold: (1) coping with complex 3D scenes, and (2) achieving persistent autonomy. These two challenges are highly coupled, and to deal with them, our research group is working to create a new set of tools for coping with the tremendous amounts of data that a mobile robot's sensors can provide. Some of the questions that we wish to pose are: Can we provide robots with a long-term autonomous existence, enabling them to deal with changes in the environment, to recover from mistakes, and to achieve life-long learning? Can we create a robot (or team of robots) that can actively and repeatedly explore a portion of the world, building and maintaining a database that can be efficiently indexed and rapidly queried, yet easily modified as the world changes? Can we develop a system in which these robots mingle effortlessly with people, merging human acquired and human annotated data with robot-acquired databases from physical sensors?
I am also part of MIT's DARPA Urban Challenge Team
Publications
Some recent papers:
E. Olson, J. Leonard, S. Teller. Fast Iterative Alignment of Pose Graphs with Poor Initial Estimates. In Proceedings of the 2006 IEEE International Conference on Robotics and Automation, May, 2006.
M. Benjamin, J. Curcio, J. Leonard, P. Newman. Navigation of Unmanned Marine Vehicles in Accordance with the Rules of the Road. In Proceedings of the IEEE International Conference on Robotics and Automation, May, 2006.
M. Walter, R. Eustice, and J. Leonard, A Provably Consistent Method for Imposing Exact Sparsity in Feature-based SLAM Information Filters, In Proceedings of the 12th International Symposium of Robotics Research (ISRR), San Francisco, CA, USA, October 2005. [details]
R. Eustice, M. Walter, and J. Leonard, Sparse Extended Information Filters: Insights into Sparsification, In Proceedings of the 2005 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), Edmonton, Alberta, Canada, August 2005. [details]
R. Eustice, H. Singh, J. Leonard, M. Walter, and R. Ballard, Visually Navigating the RMS Titanic with SLAM Information Filters, In Proceedings of Robotics: Science and Systems (RSS), Cambridge, MA, USA, June 2005. [details]
E. Olson, M. Walter, S. Teller, and J. Leonard, Single-Cluster Spectral Graph Partitioning for Robotics Applications, In Proceedings of Robotics: Science and Systems (RSS), Cambridge, MA, USA, June 2005. [details]
R. Eustice, H. Singh, J. Leonard. Exactly Sparse Delayed State Filters. In Proceedings of the 2005 IEEE International Conference on Robotics and Automation.
D. Moore, J. Leonard, D. Rus, S. Teller. Robust distributed network localization with noisy range measurements. In Proceedings of the Second ACM Conference on Embedded Networked Sensor Systems (SenSys '04). Baltimore, MD. November 3-5, 2004.
R. J. Rikoski, J. J. Leonard, P. M. Newman, and H. Schmidt. Trajectory Sonar Perception in the Ligurian Sea. In Experimental Robotics IX, 2004.
E. Olson, J. J. Leonard and S. Teller, Robust Range-only Beacon Localization, AUV 2004.
M. Bosse, P. M. Newman, J. J. Leonard, and S. Teller. SLAM in Large-scale Cyclic Environments using the Atlas Framework. International Journal of Robotics Research. (multimedia)
P. M. Newman, J. J. Leonard and R. J. Rikoski. Towards Towards Constant-Time SLAM on an Autonomous Underwater Vehicle Using Synthetic Aperture Sonar. Proceedings of the Eleventh International Symposium on Robotics Research, Sienna, Italy, October, 2003.
J. J. Leonard and P. M. Newman. Consistent, Convergent, and Constant-time SLAM. IJCAI 2003. (multimedia)
J. Leonard, R. Rikoski, P. Newman, and M. Bosse. Mapping Partially Observable Features from Multiple Uncertain Vantage Points. International Journal of Robotics Research, Volume 21, number 10, pages 943-975, October, 2002. (postscript)
J. Tardós, J. Neira, P. Newman, and J. Leonard. Robust Mapping and Localization in Indoor Environments using Sonar Data International Journal of Robotics Research, Volume 21, number 4, pages 311-330, April, 2002.
Contact Information
MIT Dept. of Mechanical Engineering 77 Mass Ave., Room 5-214 Cambridge, Massachusetts 02139 Telephone: (617) 253-5305 Fax: (617) 253-8125 Email: jleonard "at" mit "dot" edu and MIT Computer Science and Artificial Intelligence Laboratory 32 Vassar St., Room 32-332 Cambridge, Massachusetts 02139 Telephone: (617) 253-0607
