RoboGolf

Surprisingly, with the breadth of research topics in the Robotics Institute, many students go through their graduate career without ever building a robot. Thus, each year, part of graduate student orientation involves a robot-building competition. This used to run as a 'winner-organizes-next-year' event, but recently, it has become more of a burden with the continually growing number of admitted students. I revitalized this program in a few ways.

First, each year used to feature a different challenge, which made the prospect of organizing more daunting. I thought up a task that was not only fun and feasible, but I also designed and built the courses such that they will last many years and can be stored easily.

In previous years, the competitions used lego mindstorm kits. While it is certainly possible to create cool robots with those kits, you don't quite get the whole robot building experience until you are cutting through metal with a hacksaw (or at least trying your best to do so). We switched to VEX robotics kits which are a bit more advanced, increasing the challenge for our incoming graduate students.

One thing that differentiates my thinking about competition is that it does not always have to be head-to-head. In prior years, the competition was structured in tournament brackets, where two teams compete in one arena, where one is knocked out and forced to pack up early. This meant that some teams spent all day building a robot just to get knocked out in the first round. In order to make sure everyone has a positive experience, I eliminated the tournament style bracket and switched it so that teams are racing against the clock. Each team has the same amount of time to complete each hole, meaning everyone gets the same experience.

A more detailed description of the rules can be found here.

We have had a lot of success running this event for two years now. You can check out highlight reels for 2017:

and 2018:

The full results for each team can be found here. If you are interested in setting up a similar scoring system, you can find my friend Zhi's code on GitHub.