Blogs

FAROS navigates a smoke-filled and obstructed hallway (Credit: KAIST)

From Gizmag

 BEN COXWORTH   JANUARY 20, 2016

Fires in high-rises can be particularly deadly. This is partially because of the buildings' "chimney" effect, along with the fact that it's just plain difficult for firefighters to reach the flames. With that in mind, researchers at the Korea Advanced Institute of Science and Technology (KAIST) have created the flying, wall-climbing, fire-resistant FAROS quadcopter. It's designed to ascertain the source of a fire as soon as possible, along with the locations of people trapped within the building.

A new-and-improved version of a 2014 project, FAROS (Fireproof Aerial RObot System) is intended to scout the interiors of buildings before firefighters enter, allowing them to save time and minimize risk by knowing in advance where to concentrate their efforts. It can autonomously fly down hallways, guided by a 2D laser scanner, an altimeter, and an inertial measurement unit containing accelerometers and gyroscopes. These sensors also allow it to estimate its location within the building.

In situations where its flight path is partly blocked, it can turn sideways and press itself feet-first up against the wall, then use its propellers to push itself along until it clears the obstacle – it's not unlike the wall-climbing system used by Disney Research's VertiGowheeled robot.

Using a thermal-imaging camera along with a dedicated image-processing system, FAROS can detect people, plus – if things aren't too far along – it can ascertain the fire's ignition point. This data is transmitted back to firefighters.

Additionally, FAROS features an aramid outer skin, protecting its electric and mechanical components from flames. There's also a thermoelectrically-cooled air gap between that skin and the drone itself, helping to isolate it from intense heat. In lab tests, the copter withstood butane gas and ethanol aerosol flames reaching over 1,000º C (1,832º F) for more than one minute.

It can be seen in action, in the video below.

Full article here