From David Lu on ros-users
Inspired by ROSCon this year, I realized that there was quite a bit of code that our lab has developed that we had yet to push to our public repository. For links to each of the packages, check out our repository's wiki page here: http://ros.org/wiki/wu-ros-pkg
Note that new stable WashU stacks are being released on a branch of our Google Code repo http://code.google.com/p/wu-robotics/source/browse/#svn%2Fbranches%2Fstable . If you are still referencing the Sourceforge repo, that will remain up, but not updated.
A quick overview:
We have a number of simple PR2-based demos that we developed since getting our PR2 last summer. These are mainly used for showing off the robot to the general public. They include demos where the robot mimics human motions (pr2_kinect_teleop), has simple conversations (pr2_chat), looks at recognized faces (face_follower, formerly known as pr2_creeper) and swings a plastic lightsaber (pr2_sith).
There's also a tool for doing some basic tasks, i.e. controlling the projector and spine, from the command line (pr2_commandline)
We had previously released a driver/node for the B21r. Thanks to Thibault Kruse, Christina Lichtenthaeler and Mikhail Medvedev, we have not only improved this stack, but also added initial support for the atrv-jr. I have also added in a few improvements to the B21r node based on Chad Rockey's ROSCon talk on Hardware Drivers, upgrading the node to use dynamic_reconfigure and diagnostics. I've also added a teleop program we use, as well as a new dashboard. (See next section)
WashU's ROS Tools
There are a number of general ROS tools that we've developed here that others may also find useful.
First, a package for making dashboards like the PR2 and turtlebot have (generic_dashboard). It's not complete yet, but there's enough capability that I was able to make the B21r dashboard.
Second, there's a forked version of rxgraph (rxgraphplus) which allows for hierarchical organization.
We also have tools for * generating visualization markers quickly (easy_markers) * playing a bag as though it is happening now (rosbaglive) * checking how complete your manifest.xml and stack.xml files are (manifest_cleaner)
Since joint_state_publisher and simmechanics_to_urdf have been absorbed into the core ROS stacks, the versions in our repo have changed their names to *_experimental. You can also find the URDF from my ROScon talk in this stack (roscon_urdf).
Some tools for reading in MIDI files and sending around musical information can be found in the music stack. We also have a demo from one of our rotation students that implements differential dynamic programming in ROS (color_DDP).
Besides the work that makes up our actual research, here's a few things we hope to get out in the next months. * Merging urdf_python into the urdf_parser package * Further motion_capture analysis tools * New people recognition tools
Cheers, David V. Lu!!