Image credits (clockwise from top left): Google, University of Toronto, WillowGarage, Atlas Copco, University of Oxford, John Deere/iRobot, NASA/JPL/Caltech, Rutgers University

Note:  Slides and Abstracts as PDF can be found on the Talk Abstracts Page

Details

When: Friday, May 10, 2013, 8:30am--5:00pm

Where: Kongresszentrum Karlsruhe

Scope

State-of-the-art robotic systems continue to make strides in terms of their ability to operate independently for progressively longer and longer periods of time. Existing robots have demonstrated robust, autonomous operation for durations on the order of days. However, to be most useful, future robots will be required to extend this period to months, years, or more. In particular, long-term navigation capabilities will be critically important. While current navigation and mapping algorithms are able to function over substantial areas, it is still unclear how to extend these algorithms to deal with human-scale spatial and temporal dimensions. This focused, full-day workshop will concentrate on problems related to the development of long-term navigation and mapping technologies, for use in unstructured and highly dynamic environments. How can we enable consistent state estimation over very long time periods? What types of representations are most suitable for capturing both static and dynamic aspects of the environment? What role should online learning play? We invite researchers to share their experiences and insights. 

Topics of interest include, but are not limited to:
  • Persistent state estimation in dynamic environments 
  • Suitable data representations for long-term mapping
  • Place recognition over extended periods of time
  • Reasoning about environmental change
  • Lifelong learning for navigation and mapping
  • Long-term planning and exploration strategies 
  • Use of semantic information to improve performance 
  • Resource-efficient navigation methods

Speakers

The list of speakers will be updated as confirmations are received.
  • Prof. Peter Corke, Queensland University of Technology
  • Prof. Wolfram Burgard, University of Freiburg
  • Dr. Tom Duckett, University of Lincoln
  • Dr. Michael Kaess, Massachusetts Institute of Technology
  • Dr. Fabio Ramos, University of Sydney
  • Gajamohan Mohanarajah, ETH Zürich

Organizers and Contact Details

Dr. Jonathan Kelly, Robust Robotics Group, MIT
jskelly (at) csail.mit.edu

Dr. Paul FurgaleAutonomous Systems Lab, ETH Zürich
paul.furgale (at) mavt.ethz.ch

Prof. Gabe SibleyDepartment of Computer Science, George Washington University
gsibley (at) gwu.edu

Prof. Tim BarfootAutonomous Space Robotics Lab, University of Toronto
tim.barfoot (at) utoronto.ca 

Venue

The Workshop on Long-Term Autonomy is a full-day workshop of the 2013 IEEE International Conference on Robotics and AutomationKarlsruhe, May 6 - 10, 2013.