It' s been a while since my last posting, but I haven't been totally idle.

Firstly: the robot body has undergone yet another rebuild. I discovered that the wheels I used in the previous version wasn't robust enough to handle turning properly: The tires tended to get twisted off the rims. I've instead reverted to some smaller wheels with harder and wider tires. These have less grip than the others, but the robot still manages to climb over the door-sills. The robot chassis has also been completely rebuilt, since the previous one didn't feel robust enough. The new chassis consist of a box made up of interlocking beams, and feels much more rugged:



Secondly, I have received the bluetooth radio-module, and all the sensors that I ordered from Sparkfun. I've currently hooked up the radio, the compass, and a distance measurement wheel. I've also made a PC-program that communicates with the robot via bluetooth:
This program can be used to manually control the robot with a gamepad, and to display various status-information from the robot, like the current compass heading, battery voltage, etc.

So much for the fun stuff. I've learned a lot from this project, but one of the things I've learned is that a magnetic compass's are a hassle to work with.
1. The robot has to move relatively close to large speakers, which can totally confuses the compass (with up to 180 degrees).
2. The magnetic fields of the front-wheel motors are also a problem, since these motors are rotated when the front wheels are steered. This distorts the compass reading with up to ten degrees.
I'm not quite sure how to overcome these problems. It should be possible to compensate for the distortion caused by the robots own front motors, since the robot knows the angle of each of the front wheels. But problem (2) is harder. I might have to measure the magnetic field at several places in the appartment, so that the robot can compensate the measured heading with the known properties of the magnetic field at its current position.