It’s never easy to do what Kevin Nguyen, Kyle Yotter and other members of Dunwoody Robotic Snowplow Club did: build a device that could clear snow.
This winter, the task was made more difficult by the almost complete lack of snow.
For the 14th annual Autonomous Snowplow Competition, organizers had to make snow, then shovel it onto the Dunwoody College of Technology parking lot so the robots could attempt to plow it away.
“Those students have built autonomous snowplows designed so that when it snows, you can literally open up your garage, hit a button and it will go out and plow your driveway, sidewalk or whatever you want,” said E.J. Daigle, dean of robotics and manufacturing.
More than 80 students from the Upper Midwest took part in the competition, which was hosted by the Institute of Navigation. It featured a T-shaped field designed to simulate a sidewalk that intersects a driveway. The robots start in what would be the garage. The robots have to plow as much snow as they can while avoiding obstacles — bright orange cones that represent mailboxes, parking meters or even a grandma, Daigle said. Snow removed is measured and penalties are deducted for hitting obstacles.
Some teams used GPS, radar, sensors and magnetics to navigate their robots. While some of the navigation systems were sophisticated, many of the teams struggled because they only got to practice on the morning of the event, so they calibrated their machines to a shady environment. By the time the competition started, the sun was directly overhead. (One clever team put sunglasses on their robot’s camera lens to decrease the sun glare.)

Open to the public, the competition typically attracts more than 100 spectators and is livestreamed online via drones with cameras. Daigle said it gives students a chance to compete and gives them the opportunity to shine in front of companies interested in developing autonomous snowplows.
When it was time for Nguyen and Yotter’s robot, Ice Hawk, to perform, the duo placed QR codes on boards on the sides of the snowy field. A webcam on their plow was designed to sense the codes, to help it measure exactly where it is on the pavement.