Meeting Summary
from: 3Feb99 at 10:00am 
with: Chris, Ben, Ian, Chanop, David



Agenda
  • Status on individual activities



  • Ian continues to make progress on the framegrabber device driver and is now getting color images into memory (or at least 1/4 of a color image).

    Ben has the camera calibration system working with his own targets design.  He has been working to eliminate outlining points which perturb the line fitting algoritm he is using.  A more robust method will evenutally be needed but for now poorly fit edges can be ignored with still enough edges remaining to perform the calibration.  The next step will be to estimate the distance to some points and try pointing the zoom camera at them.  On Thursday at 2:00pm Ben will give a short seminar on his work this summer.

    Chris has new, much faster backpropogation code working, thanks Jon Baxter.  He has been studying the value functions used by various reinforcement learning systems and looking for a good model to use.  This has lead to an investigation of optimal control (where the cost function is the negative of the value function used by reinforcement learning.)

    Chanop has been working on the digitizer operational under Linux.  He will continue to revise his master's research proposal by expanding the detail and incorporating references to the material he has been reading.

    David continues to chase up components and outstanding work (endcap, battery charger, serial cable, etc.)  The final component, a fiber optic media converter has been ordered.  We hope everything from here on out will be small items.  David seems to have the memory mapping worked out for the PCI carrier boards so that the IP modules can be probed.  Next step is to get Bob Edwards driver working.  David still needs to finish the wiring database.  The project schedule has been updated with comments recieved so far--it will probably need a revision once all the honors students work out their individual projects.

    Ali is still missing in action with our amplifier designs.  Hopefully he will turn up this week.


    Kambara Project at the ANU Robotic Systems Laboratory
    By David Wettergreen<dsw@syseng.anu.edu.au>
    Last modified: