Robot 1

More pictures here.

This was my first robot, other than one I made out of wood when I was about 8 years old.  I remember being disappointed when I realized the wood one wasn't going to actually do anything.

Well this one actually does something, but not much given the poor design of the Tamiya robot tracks I bought.  I'd recommend you stay away from them unless you're planning to drive only on smooth surfaces; the rubber tracks come off all the time on my carpet at home, so this bot is now in mothballs.

Here are the features:

• Remote control using laptop keyboard and Xbee antennas.
• Arduino brain.
• Parallax ping sensor mounted on custom head to allow directional distance measurements.
• Generate tones for feedback and fun.
• Cable free operation.

I started this in the early summer of 2010, and considered it finished about a month or two later.

It pretty much just drives around and tries not to bump into stuff when in autonomous mode, and remote mode allows for control of vehicle movement, head positioning, and measurement taking using the ping sensor.

The robot uses an Arduino controller as the brain, the motor controller shield from Adafruit, and a Tamiya gearbox, base, and tracks.  I made the circuit platform and head out of plexiglass.

I wrote a program in C# to send commands to the robot over the wireless COM connection provided by the Xbee.  Steering is identical to what you find in a first person shooter game, the buttons ADWS control forward/back/right turn/left turn.  Nope, no strafing mode with those tracks, but those omni wheels look interesting if I ever decide to get into wheeled robots.

Other buttons recognized by the bot are "Q" to engage autonomous mode, "L" to engage look mode, "M" to measure via ping the current distance to an obstruction, if any.  Those last two were extremely helpful for me to understand the limitations of ping sensors generally, as there are wildly bad measurements when you are pointing at something acoustically reflective, like a guitar or a hard countertop.  Testing in later bots got me feeling that Sharp IR sensors are the way to go, much more reliable in a standard indoor environment.