CMUQuadrotor: Autonomous Exploration From Above
NOVEMBER 21, 2010 | MIKE ORNSTEIN
Mapping and exploring unknown and potentially hazardous areas is a task often delegated to robots. From bomb-sniffers to undersea explorers, robotics is often the only way to keep humans out of harm's way. Exploring these environments is typically done via remote control, creating what are essentially robotic tools ultimately controlled by human brains. What happens when an environment is not only too hazardous for humans, but prevents remote control from being a practical solution?
Underground mines with no signal reception, disaster areas with too much interference, and even foreign planets all match this description, yet they must be explored. CMUQuadrotor aims to solve this problem.
The CMUQuadrotor project, funded in part by a Spring 2010 SURG from the CMU Undergraduate Research Office, is an attempt to develop a stable aerial platform capable of autonomously mapping unknown environments in both GPS-enabled and GPS-denied situations. By combining an array of sensors ranging from an inertial measurement unit (IMU) to sonar rangefinders and eventually to vision or light detection and ranging (LIDAR), the project aims to build reliable maps of uneven, unknown environments so that humans or other robots can safely enter them. The platform, the Quadrotor, also hopes to be available for aerial photography and videography of campus events like sports games, carnival events, and buggy races.