Warning: include_once(scripts/rss.php): failed to open stream: No such file or directory in /userfiles/CS/public_html/eclab/projects/robots/arc/local/config.php on line 92

Warning: include_once(): Failed opening 'scripts/rss.php' for inclusion (include_path='.:/usr/share/pear:/usr/share/php') in /userfiles/CS/public_html/eclab/projects/robots/arc/local/config.php on line 92
Applied Robotics Club : Main - Pioneers browse

Pioneers


Projects

Getting Started Guide

We really need a guide to help people get started working with the Pioneers. Needed topics include:

  • Installing Player/Stage
    • Linux
    • OS X
    • Windows
  • Using the Pioneer laptops
    • Logging in
    • Where to save projects
    • Paths to commonly needed include files and libraries
  • Hello World of Robots
    • Point out necessary elements in main *.cc file
    • How to build
    • How to run in simulator
    • how to run on actual robot
  • How to use GMU Player library

Convert existing work to Player 2.0

There's a new API for Player. We'll need to re-write some of our code to get it working with this new API.

Monte Carlo Localization

One of the challenges in robotics is answering the question, "where am I?" With Monte Carlo localization (also known as a particle filter), a map of the environment is known ahead of time. The robot may, or may not know where it starts.

  • Implement MCL, like what was developed for Trinity last year
  • Add KLD sampling
  • Kidnapped robot problem

Occupancy Grids

Occupancy Grids are one method of map building. Discretize the world -- carve it up into small squares, and mark each square as occupied, or unoccupied. More specifically, assign a probability to each square, to indicate the probability that it is occupied.

  • Implement wall follower
    • hey, we need some exploration algorithm...
  • Implement occupancy grids algorithm
    • Should work very well in Player/Stage
    • Should fail in the real world

SLAM

The localization problem becomes much harder when you don't have a map ahead of time. In this case, the robot must perform simultaneous localization and mapping (SLAM).

  • Implement Grid-based FastSLAM
    • Combination of MCL and Occupancy grids