It comes equipped with a plow, several engines and a laser. Don’t worry, the laser is just for guidance.
M.A.C.S., a completely autonomous snow-plowing robot, helped a team of electrical engineering undergraduate students from Ohio University clean up in St. Paul, Minn. Led by its laser through a series of beacons, it cleared two one-meter paths of snow better than any of the other robo-plows in the competition at the city’s Winter Carnival.
“It feels awesome to do something like this,” says Samantha Craig, Team M.A.C.S. member alongside fellow juniors Matt Miltner and Derek Fulk. “It was a lot of hard work and late hours, but it was so much fun. You learn so much more working it out than just sitting in the classroom.”
Organized by the Institute of Navigation, the first robotic snowplow competition brought together teams from six universities to develop the year’s best fully-autonomous snow plow robot. ION, responsible for the annual autonomous lawnmower competition, took a turn for building a better machine for the winter season.
“The challenge we faced here was quite difficult from the lawnmower,” says Wouter Pelgrum, team advisor with Frank Van Graas. “Just having that plow go in a straight line is far from trivial.”
A plow needs to be able to hold up on a slick surface and move heavy snow, and can’t rely on GPS while surrounded by buildings in an urban landscape, both tough for a small machine.
The students, from Ohio University’s Russ College of Engineering and Technology, had to come at the problem from a different angle. It took the entire summer and into the winter, with 80-hour weeks of building and testing.
“We worked with a new idea, starting with the frame of a search-and-rescue robot from a past project,” says Craig.
They refitted the robot with four 5 HP electric motors used for wheelchairs and weighed down the machine to about 500 lbs. to help give it more traction with which to use that power. They attached a straight plow to keep from fighting against the weight of the snow on the blade.
“With plowing, you need weight and you need power,” says Pelgrum. “We went full throttle and did what we needed to win.”
Armed with those motors, M.A.C.S. had plenty of muscle, but still needed a brain – or an eye, more appropriately. The machine was equipped with a 360-degree scanning laser, searching its surroundings at a level of 5Hz to find sections of PVC set as beacons around the course. The robot was named for its laser sight: “Monocular Autonomously Controlled Snowplow,” though that came after a good deal of discussion among the student team.
“I don’t want to say it was the hardest part, but it was a pretty big deal,” says Craig.
The laser reflected to the robot measured its distances to the landmarks, relying on an on-board gyroscope. Once M.A.C.S. had placed itself with that data, it could use algorithms to determine its heading and what motion needed to be performed next.
Being able to constantly read and orient itself to beacons, M.A.C.S. was able to clear snow off the track in laser-sharp straight lines, pushing forward, rounding two corners and then coming back for a second pass all within the competition’s time limit, completely without direction from any of the humans who worked on it.
“A lot of people don’t know exactly what autonomous means for robots,” says Craig. “It has to operate without any human interaction. It’s a machine where you just push the start button and it goes.”
And it might go on to develop into a machine that makes work for snow plow professionals a little easier and maybe safer, according to Pelgrum.
“We’ve been brainstorming – you could combine a snowblower with this and do automated snow removal from rooftops,” says Pelgrum. “You could use this in more difficult-to-reach areas, or anywhere you would have to plow extremely often, a robot could be useful.”
The team is using the $2,500 award for a first place finish and the additional $500 for top design toward the students’ education and development of next year’s model. The advisers are hoping to get some comments and ideas from snow plow professionals and come back next year with a machine that shows a glimpse even farther into what a robot can do for snow removal.
Want to see M.A.C.S. in action? Watch its journey from workroom to snowplow victory right here!