As a part of 2.72 (Elements of Mechanical Design) in fall 2004, my team had to redesign a stock hexapod robot (a picture of it is below) to compete in a tug-of-war contest with other robots. Our idea was to try and maximize the number of servos acting at once, and thus maximize the available pulling force, by having the robot rotate about its center and thus "reel" the tug-of-war line in, instead of pulling the rope in a linearly as all of the other designs did. So, we made a hexagonal robot that rotated (we had 6 servos to work with). The robot was semi-autonomous, controlled by a BASIC Stamp and a remote control which could select from the various routines. So, along with helping design and construct the physical robot, I was also responsible for programming the various walks, we called the main two sequences the ripple walk and the starfish walk(my code is available upon request). The ripple walk had six states: resetting, just touched down and pulling back, pullback state one, pullback state two, pullback state three, and the final pullback state. Each leg went sequentially through each of the states, resulting in one leg being in each of the six states at any one time. The starfish walk consisted of a reset function, moving all of the legs all the way to one side, and then a pull function, where every servo moved the same direction at the same time (obviously this is not how real starfish move, but "unison walk" isn't as catchy a name).

After a mock competition, we realized that we needed a guard to stop the tug-of-war string from touching the electronics mounted on the top plate, and so that was attached to the top, underneath the spool. In the final competition, though, a few design flaws led the robot to break. First, the legs were attached to the servo horns poorly, and a few of the horns broke. Also, there were a number of optimizations concerning weight and general mechanism design that kept us from exerting the maximum amount of force. For the final round of competition, we had neither the material nor time to make the robot a winner, so we adapted the robot to use its original legs from the hexapod kit, just so we could compete. The robot performed worse in the final round, but that was as expected. I got the first and last two photos on this page from my teammate WeyWey.