George Westwater

This weekend was mostly filled with the smell of solder, piles of wires and polymorph, My goal this weekend (which I fell a little short of) was to have R2-A6's (color changing paint makes him also a R2-A3) vision sensors prototyped. Unfortunately the chips were stacked against me, sparkfun did not get my order to me when I had hoped instead it will be here Monday so I only had 1/2 the sensors I wanted to use.

Next when I was learning / experimenting with each of the sensors and the LED screen I noticed my LED screen had some damage to it (just not my day). However despite all of that I was able to get the a few components "working".

For those of you whom do not know my sensor plan for R2-A6 is to have the very top of his dome extend vertically by about 3-4 inches which will expose 4 sensors configured in a square (see below). These will provide the room mapping and motion detecting functions for R2. This "map" will be fed into a Raspberry PI using the mesh communication protocol I have been developing on top of ZigBee (I might change to Wixel to increase flexibility). The Rasberry PI will then help pilot R2-R6 in autonomous mode. More information on the protocol and control algorithms to follow. And yes before you mention it I looked at ROS and found that it did not mesh (as in mesh networking) the way I wanted and it had some limitations on clustering and dynamic module inclusion which would not work for me. If you are interested in ROS you can get it from here. I am however spending some cycles looking into OpenCV. The jury is still out on this one.

Anyone off the rant and back onto the weekend progress.

Continuous Rotation Servo Control:
I was able to get this working almost to where I am almost happy with it. The accuracy is close (not close enough yet). I may have to add some optical encoders or stop switches to ensure I can "reset" back to position 0.

PING Distance Sensor:
This thing works very well, actually much better than I expected however the range is not exactly what I would have hoped for so I may migrate to a XL-MaxSonar or at least buy one to test it out.

PIR Motion Sensor:
Yes this works, however it is FAR to sensitive so I am working out some ways to fix that by modulating the duration of the power,

JPeg Camera:
Has not arrived

IR Distance Sensor:
Has not arrived

GPS:
It is really remarkable how well this works, however the embedded antenna is to small and it has a lot of trouble picking up anything from my basement.

Mapping Functions:
The software so far is turning out really well, however it is really hard to cram all of this into about 15k which is all I have to work with. However so far I am still very happy with it. I am currently outputting the room map to a small LCD screen, and that still needs work.

I also build a simple servo and PING sensor harness to help me test and collect data for the room mapping function you can see a picture below.