Hi Guys and Girls! Long time no see!
There is a competition in my country which is called Hungarians On Mars, official site. It is held once in a year, and it’s about robotics challenges. As always, we attended.
This years challenge was to make a two wheeled robot and capture points by dropping magnet onto them while other three challengers try to do the same thing.
My friend bought a Tenda W311mi and when he tried to use with his BeagleBone he got stucked. So I thought that I will help him to find a solution
Robots self testing
As we build more complex robots we need more feedback from them to know what state they are in. In my case I want to know some temperature information about my rover. These things are the motors, the outside environment, and the battery temperature. For this purpose, I want to create an Arduino thermal sensor. Later it will be only an Atmel chip with the sensors but for prototyping and developing the Arduino is the best solution.
The central computer on the Rovers will be Beaglebone. So after I create the thermal sensor, I need to make the communication between these devices. If that is done, the only thing is left to make the ROS package. I have chosen ROS because in the near future I want to make more sensors and ROS is ideal for information distributing.
With the OS of Debian new opportunities have come. The Angstrom distribution did not meet our requirements because conman and ROS was not supported well yet, while Ubuntu didn’t have PWM, so that’s how Debian came into picture. The idea of connecting to the BeagleBone directly is better then connecting through some middle layer (like a router). To be this simple we need to config an acces point on the BeagleBone.
Yesterday we went to our evil genius friend to ask him to make a new rover for us. Instead of making a new one we ended up reviving an old rover.
This rover was our robot in the competition of Magyarok a Marson 2011 (Hungarian on Mars). The competition was similar to a capture the flag game, with four rovers on the field at the same time.
We weren’t that experienced that time (and our budge was low) so the project failed. We were good in the first round but unfortunately at the second round some smoke has been made and the motorcontroller went to the robo-heaven. Also we smoked an 25A car fuse, yeah something was very bad in our design. We didn’t have the automatic control in time so we used the manual version. In this case we had to use 15 sec delay between two command to follow the rules. We started the blog roughly after two months to the competition. Here is a picture of the rover at that time.