After completing the "technology upgrade" of the LPCR to include the Arduino, I think I have to make a note about the different strategies of the two platforms.

Linux is a powerful multitasking system. It has powerful features such as multiple processes, shared memory, multiple threads within processes, lots of memory, and formalized disk access. As a multitasking system, many programs compete for resources such as CPU time, memory, and I/O.

If you are at the Boston USENIX Conference, I am giving a talk about the Linux PC Robot in the Plenary session at 4:00~5:30. Neatly sandwiched between break and happy hour!

Update: 6/23/2012
Well, it seems that people had a good time. The G*&^&D D*%$#M wireless has again been a problem. How do you show technology that needs wireless communications in a function hall in which a hundred android users are browsing Amazon for books!

Really...seriously...It works!!! LOL. Funny story, I was able to do a mini-demo to one of the USENIX sys admins, shout out to Anthony!

Last year I started a redesign of the Linux PC robot to eliminate the PS/2 mouse encoder system. Between work and family, it has been difficult to get the time to get it done.

I've been trying to get a good remote viewer app to use a web cam in conjunction with either a control panel or a joy stick to control the robot remote.

I have searched high and low for a low latency video streaming system. I could not find one. Everything I tried from gstreamer to to ffmpeg and others, all introduced a latency I could not afford. I ended up having to write my own.

After a rather lengthy break, I'm at it again and working on the arduino motor I/O component.

I have Arduino 1.0 compiling the code after some changes and all looks good.

To recap:
I am replacing the K8055 data acquisition board and PS/2 ball mouse board with an Arduino.

The first step is to implement the old methodology in the Arduino. The second step would be to move more of the motor control and basic functionality into the Arduino as well.

With a full time job and a family, its hard to get time, but I did some work over the weekend. I have a preliminary schematic of the interface circuitry and I've done some tests on the system. It looks like the CPU is fast enough to process the encoder interrupts for each wheel at full speed under no-load.

I'll try to get some hard numbers posted soon.