MathWorks gave eight teams of contestants golden tickets to Artisan’s Asylum, one of the largest makerspaces in the U.S., to conceive and construct—within just a single week—autonomous robots that would wage battle at the Grand Finale of the Cambridge Science Festival. The teams crafted their robots with a standard kit for a skeleton, an Arduino for brains, Simulink for its mind, and custom armor forged with a computer-controlled plasma cutter. The team that rose above the rest won an autonomous quadcopter with an Arduino, so they can take the lessons they learned on four wheels and take to the skies with four propellers.
Rules of engagement
Metal clashed against metal, rubber teared on wood, until only one robot remained in the ring. These robots were streamlined sumo warriors gorged full on lithium ion batteries. No heavier than 10 pounds, each robot would have to find its opponent and cast it out of a 12 foot diameter ring. For the safety of the crowd and to ensure that each team would go home with a robot in one piece, many weapons were forbidden. No fire, no lasers, no spinning blades … the list kept going because you can never underestimate the creativity of an engineer. With most weapons banned, the teams had to rely on the most powerful advantage of all: embedded intelligence.
All the better to see you with
Each robot had a continuous out-of-body experience as all robots shared an eye in the sky peering down upon them from high above the ring. Using Computer Vision System Toolbox, a web camera connected to a laptop tracked red and green placards velcroed to the backs of the robots. The placards had two patches of color on them to distinguish between the front and back of a robot, so the camera system could track each robot’s position and orientation. This data was beamed wirelessly to both robots so each robot knew where it was, which way it was facing, and where its opponent was. The robots also had infrared range finders facing forward to know when to go to ramming speed and line sensors facing downward to detect the edge of the ring.