Robomagellan 2012 Build Diary - Day 43

[eric gregori]

Days 1 - 7

http://buildsmartrobots.ning.com/profiles/blogs/robomagellan-2012-b...

Days 8 - 16

http://buildsmartrobots.ning.com/profiles/blogs/robomagellan-2012-b...

Days 17 - 27

http://buildsmartrobots.ning.com/profiles/blogs/robomagellan-2012-b...

Days 28 - 34

http://buildsmartrobots.ning.com/profiles/blogs/robomagellan-2012-b...

Days 35 - 41

http://buildsmartrobots.ning.com/profiles/blogs/robomagellan-2012-b...

Day 42

http://buildsmartrobots.ning.com/profiles/blogs/robomagellan-2012-b...

Today was a big day for the robot.  It started at 5AM.  I tested the ultrasonic range finders, the compass, and the GPS.  All those subsystems worked fine.  

I did find an interesting problem.  I had originally routed an analog signals from one of the ultrasonic sensors to the  to P1.3 on the TI Launchpad forgetting about the pullup resistor on the Launchpad.  The TI Launchpad has a pullup on pin P1.3.   

I halted testing during the day because it was too hot.  Inside I continued work on the Executive portion of the code.

At about 5PM I continued testing, doing the first closed loop compass run.  For this test, I told the robot to track a compass bearing of 0 degrees (go north).  The robot worked perfectly.  I did note that the "half speed" was still too fast.  I will be decreasing the half speed setting tomorrow.

No pictures today.  I had too much work to do to take pictures.

I did make a big decision today.  My original plan was to use the Raspberry PI for the vision subsystem.  I just don't have the time to do that project.  I am going to go with a backup plan and use an Android phone for the vision system along with an application I wrote a few years ago (Colorbot): 

https://play.google.com/store/apps/details?id=colorbot.emgrobotics....

Using an Android phone and this vision app will significantly reduce my workload.  Tomorrow I will be able to concentrate on the Executive functionality and testing instead of the vision system.  I do plan on continuing work on the Raspberry PI solution.  I plan on porting my ChumbyCV computer vision framework to the Raspberry PI as soon as the RoboMagellan project is done.  Look for a Raspberry PI+webcam based line follower in the very near future.

http://code.google.com/p/chumbycv/

Originally designed for smaller robots, I will re-task the Colorbot Android application for Robomagellan.

Colorbot was written back in the Android Eclair days.  Back in my day Android phones did not correctly support rotating the camera image :)

Colorbot was written entirely in Java.  It communicates with the outside world using light sensors attached to the screen.

[eric gregori]

Motor Processor Functions

Closed loop speed control - implemented/tested

Motion state machine - implemented/tested

Ultrasound Control - implemented/tested

Communication Watchdog - implemented/tested

Ultrasound Reverse - NOT implemented

MSP430: Program loaded. Code Size - Text: 2314 bytes Data: 64 bytes