Every year, U of T’s second-year engineering science students gather for their annual robotics competition. This year’s took place in March at the UTS gymnasium, and featured robots that shot tennis balls, detected mines, and waited tables.
Each robot must operate unaided, and has to be designed to a set of specifications presented to the students at the beginning of the course. The tennis-playing machine had to navigate to three specific positions on one side of a tennis court while firing tennis balls over the net at regular intervals. The mine-detecting machine had to navigate and identify metal plates without bumping into them. The waiter machine needed to navigate across a floor, find metal plates and put pop cans on them.
Before beginning construction, students ran computer simulations of their robot designs. But computer simulations can’t prepare a team for everything—during the competition, one of the tennis-playing machines began to spin around during its trial run, firing tennis balls at spectators.
The course, currently headed by U of T engineering professor Reza Emami, is one of the most memorable experiences for engineering science students at U of T. Course alumni reflect on the experience as a transition between designing machines in theory and actually building them. Most agree their final robot was much different than originally expected.
One of the top entries in the waiter competition was the Camarero, built by Ivana Konvalinka, Ashley Taylor and Warren Uniewski. The students attributed their success to keeping their original design simple.
Some robots had complex sub-systems and their success depended on whether or not all of their parts were working. This often translated into frustration and time lost dealing with a problem. The Camarero’s team only had to make minor changes. When their pop can dispenser wasn’t working, they used duct tape to modify the shape of their dispenser chute. Similarly, when the robot’s wheels weren’t making enough contact with the ground, the team increased the diameter with duct tape. The only major change they made was to trade control for more power by replacing the motor.
Another factor in Camarero’s success was its use of centered light sensors to keep itself above navigation grid lines on the competition floor. This allowed the robot to actively compensate for any drift in its drive system rather than just assuming it was going straight.
