Meeting Notes from 8/3 and To Do List as of Morning 8/6

Meeting Notes

• Discussed next steps in trying to get communication working with IRC5
  • Try TCP (Russell's code from 109 includes both TCP and UDP. I don't know if he was successful with both.) If it works, the problem is either ABB's UDP setup of my UDP listener.
  • Try my code/config on the IRC5 and IRB-140 downstairs. If they work, then the problem is with my ABB hardware.
• Discussed SWRI's progress and whether their work is applicable. They seem to have made progress working with ABB hardware, but their work is still incomplete. We have a formal agreement with them, so Wyatt sent them an email asking what their status is.
• Need to get localization/navigation working
  • Set LIDAR TF to match Abby's config
  • Check gyro
  • Tune Eric's steering algorithm
  • Try Eric's premade "tracks" for navigation through 210/Glennan 2
• Want to get videos of driving by next week
  • Grab a camera from Freedman center of Thinkbox
  • Get some videos of autonomous driving
  • Post to YouTube (get account access)
  • Find out who controls mobilerobots.case.edu
  • Make Abby a homepage
• Need to get gripper manufactured. Arkady can do it for me, but I'll have to order some backing material for him.
• We'll want a second PC to handle Kinect data
  • There's a small PC in the lab, left over from (probably) a senior project. It might suffice. It has a DC-DC supply, but it might only be a 12V one.
  • If not, we can spec out another PC for Abby. This would require another DC-DC supply, mobo, processor, case, and RAM. It would probably stack on top of the existing PC.
• Greg and Miles have found that the model of Abby is doing a lot of unnecessary collision calculations because it is made of so many independent (but fixed) pieces. I will try to combine all of the fixed parts into a single STL model for them. Unfortunately, this might have less accurate inertial characteristics.

To Do (no 0rder)

• Test RRI with TCP
• Glue encoder wires in (hot glue) because I'm derpy and didn't use modular connectors (revised version does)
• Mount Kinect
  • Fabricate Kinect standoffs
  • Fabricate Kinect mounting plate (need to track down or approximate a sheet metal braek for this one)
• Test gyro
• Test RRI code on IRB-140
• Make simplified STLs for Greg
• Driving video
• Demo video
  • Driving with arm motion
  • Driving with picking and placing of standard carrier
• Characterize battery performance
  • Drive train only
  • arm in motion
  • arm static
• Come up with arm/Kinect calibration procedure
• Set up shelf for Abby to grab from
• Design standard carrier
• Give Abby a holding shelf/bin/something
• Add separate heartbeat signal to remote/e-stop schematics
• Figure out why the cRIO voltage is reading 0 volts (this does not affect performance)
• New e-stop schematic
• New e-stop layout
• New e-stop remote layout
• Make PC auto-boot on power up (requires connection of one cable between PSU and mobo)
• Figure out why I can't connect to ROS remotely
• Wire up solenoid (requires some switching transistors and a way to interface with the PC, possible an arduino or something)
• New cRIO breakout layout
• Mount rear Kinect
• RRI->ROS node
  • Get basic RRI comms
  • Pick data to send back and forth
  • Write RAPID code and config files
  • Write ROS node and message types (if practical, be SWRI-compatible)