Date: Wed, 8 Feb 2006 01:01:35 -0600 (CST)
Subject: DPRG
X-UID: 133
Content-Type: IMAGE/JPEG; name="img2668.jpg"
Content-Type: IMAGE/JPEG; name="img2671.jpg"
Content-Type: IMAGE/JPEG; name="img2694.jpg"

The robot visited the DPRG (Dallas Personal Robotics Group) Tuesday 
evening. It was a lot of fun. The atmosphere was a bit of the late 1970s 
and early 1980s super nerd-dom. Think of the Real Genius movie except 
without the physics lab, high power lasers, and thinly veiled CalTech 
campus. Unfortunately, this more liberal technology culture has largely 
been lost. Everything is all business now.

David Anderson brought JBot which is far more impressive in person than in 
the videos on the web. It is a very mature and robust machine, durable and 
able to handle unstructured environments. It makes an impression to see a 
robot work that well and reliably. David has confidence that his robot 
will work. And it does.

JBot uses a commercial IMU (inertial measurement unit) which itself is 
very interesting. The robot really shows off what a good IMU can do for a 
robot (...or precision ordnance i.e. smart bombs). The combination of IMU 
and forward facing faceted sonar is very robust. It works extremely well.

My robot did make an impression due to it's size, power, and weight. It 
was the largest robot there with a surplus of power from the drill motors. 
People also found the front suspension interesting because it is a 
homemade design.

I brought the laptop, booted up the robot, ssh'ed in and gave commands to 
spin up the drive motors. I was confused as the left rear wheel ran in 
reverse. Another person there pointed out what happened. The left wheel is 
the mirror image of the right. Clockwise becomes counter-clockwise.
Dummkopf!

Well, I think I know how I'm going to proceed now with software 
development. Basically, I'm not going to change the methodology of the 
last year. I'm going to continue mathematics and software development just 
like I was working on hardware.

For instance, the first thing I need to do is get a Kalman filter for the 
piezo gyro, IR distance sensors, and potentiometer readings. It is faster 
to understand a result and how to apply it rather than derive or prove it.

This is the shortcut. I'm going to prioritize understanding a result and 
application rather than derivation and proof. It is analogous to buying 
off-the-shelf rather than building in engineering. I wish I had more time 
to build and derive everything. But I don't.