INNOVATIONS THIS YEAR

1 – The holonomic drive system. 

While, it may not be the best drive system for this year’s competition, it is a significantly challenging system to implement.  The programming students really do understand how it works and they learned some interesting things while programming it.  First, they had each been exposed to Vectors in their math classes, but this is a unique practical application of using Vectors in the real world.  While doing this they also learned how to re-orient the joystick through software using Vector rotation.  Game strategy is effected by the selection of drive system and was new territory for the students as well.

The following article written by Rich LeGrand and published in the Circuit Cellar magazine was a great reference to our programming mentors and student team. One can purchase a copy on-line for $1.50.

Closed-Loop Motion Control for Mobile Robotics
Rich LeGrand
Rich recently came up with two closed-loop drive train designs for mobile robots. All it took were some inexpensive permanent magnet motors and a simple feedback scheme. In this article, he covers everything from PID control and tuning to trajectory generation and operational space control for two robot bases. He also explains the software. p. 34
Keywords: Robotics, control, closed loop, PID, back EMF, C++, kinematics, holonomic

The drive wheels that we selected for our base are the AndyMark.biz plastic omni-wheels. These are a second generation wheel developed based upon the user reports by FIRST teams using the AM aluminum Trick wheels in holonomic bases in the 2005 FRC.

2 – A State Machine was implemented in this years code. 

A state machine consists of discrete operations that contain the capability to move from one state to the next state automatically.  The advantage of a state machine is that each of the individual states isn’t implemented to solve the entire problem, but rather to solve small parts of the problem.  They are then strung together to provide a complete solution.

For example, this years robot has a Move, Rotate, Find, Fire states.  Taken individually none of the states are very useful, but string them together with the capability to either be told or to determine the next state it can solve a complex problem.  The overall problem may consist of shooting a ball through the goal.  The state machine would take care of the individual steps responsible for achieving that. 

One solution may be:

MOVE Right 10 Feet, MOVE Forward 10 Feet, FIND the target, FIRE the ball.

The big advantage of a state machine is that if the rules for any single state changes we can adapt that state and still maintain the overall solution with minimum impact to the overall system.  Let’s say they replace the GREEN target with an Infrared Emitter and gave us sensors.  We could pull the FIND state out, change it to see the infrared emitter and press on without having to change the MOVE, ROTATE, and FIRE states.  Also, if we add sensors to the robot we can add states to utilize those sensors.  Let’s say we develop a piece of hardware that can detect robots and differentiate alliance bots from non-alliance bots.  We could then add states to play better defense (ie FIND_ALLIANCE_BOT state)

It’s the precursor to intelligent machines that can be developed to operate autonomously for long periods of time by utilizing small, reliable and replaceable chucks of code provide systems flexibility that can be eerily life-like.

Lone Star Regional Robot Picture Database<-- click to load the picture database

Championship - Archimedes Team List <-- click to view database