Towers of Hanoi - Senior Design 1 Project



There is a course at Clemson every electrical engineering major must take before graduating called Senior Design 1 (and 2).  When I took senior design 1 our group of 5 students was assigned the project Tower of Hanoi.  The traditional Tower of Hanoi game is a mathematical puzzle that consists of a stack of disks and three towers or rods to stack the disks on.  The goal of the game is to transfer the stack of rings from one tower to another to the same configuration as it started.  Wikipedia has a great description of the rules:  

Only one disk can be moved at a time.
Each move consists of taking the upper disk from one of the stacks and placing it on top of another stack or on an empty rod.
No larger disk may be placed on top of a smaller disk.

We had to make an autonomous system that could finish this game with a stack of 5 rings. 

Our final project could finish the game in 1 minute 37 seconds, which is about 3 seconds per move.  Our project was the fastest project out of all 10 teams on the tower of hanoi project, with the second fastest team being 3 minutes 38 seconds and the average time of all the teams being around 7 minutes.  

Parts list



There were no space constraints to our project, so we made the base of it using 2x4's approximately 3' by 2'.  The requirements of the course were to use the Quanser Q4 board (the big green thing in the pics) and a DC motor with an encoder that they provided us, but beyond that everything was up to us.  

From the beginning we knew we wanted the game board to slide to the left and right.  This system allows the part that has to move the furthest (the game board) be the lightest piece and driven by the heaviest motor (the provided Tohoku DC motor).  Our vertical movement was driven by a stepper motor, then our gripper was driven by a cheap servo motor.  

Initially our design used a lead screw for the vertical movement.  We designed it this way for maximum strength of lifting the rings.  However, during testing we found that the Quanser board had some drift when moving the stepper motor and the dozens of rotations it needed to spin the lead screw for the full 5" of movement would result in 0.5" of drift and we would miss the ring.  The lead screw system also wasn't very fast, taking 3-5 seconds to move from the top to a ring position.




I really can't speak about the software - I never had to touch it!  I was very lucky with our group assignments and was placed with a group that had some very competent and hard-working computer engineers, and after I designed/built the hardware of the project I told them I didn't want to have to touch the code at all.