US Navy firefighting robot still needs its sea legs

SAFFiR is a slow, lumbering humanoid but it could work with shipboard drones to find and fight fires

The U.S. Navy's Shipboard Autonomous Firefighting Robot (SAFFiR) undergoes testing in rain gear aboard a decommissioned Navy vessel. It can manipulate doors and fire hoses and could work with drones and human firefighters in the future.

The U.S. Navy's Shipboard Autonomous Firefighting Robot (SAFFiR) undergoes testing in rain gear aboard a decommissioned Navy vessel. It can manipulate doors and fire hoses and could work with drones and human firefighters in the future.

The U.S. Navy is experimenting with a robot designed to help fight fires on ships, but it's not nearly fast enough to be effective yet.

The Shipboard Autonomous Firefighting Robot (SAFFiR) is a 177 centimeter-tall, two-legged humanoid that can walk on uneven surfaces and manipulate doors and fire hoses.

While it's currently remote controlled, slow and can be damaged by fire and water, SAFFiR could get an extended awareness of its environment thanks to a drone that can fly through the cramped interiors of ships.

Since military vessels carry ordnance, there's heightened urgency to extinguish fires. Developed by engineers at Virginia Tech and backed by the Office of Naval Research (ONR), SAFFiR is equipped with stereoscopic thermal imaging cameras and a rotating LIDAR (light detection and ranging) laser sensor that allow it to navigate through thick smoke.

In a series of tests aboard a decommissioned vessel in November, SAFFiR was able to use thermal imaging to find overheated equipment and handled a hose to extinguish a small fire, the Navy said Wednesday in a release.

A bulky, 65 kilogram machine that's not unlike the U.S. military's humanoid rescue robot Atlas, SAFFiR requires a safety tether for now even though it has a whole-body momentum control system to help keep its balance on unstable surfaces. It has the ability to operate independently, but it's being remotely operated by engineers in current tests.

The drone developed by Carnegie Mellon University and spinoff Sensible Machines, however, does operate autonomously and performs all its computations onboard. It's able to navigate through doors and smoke using a depth camera, which is similar to a Kinect 3D game controller, while using a downward-facing optical flow camera to keep track of its own motion. It's 58 cm wide and has four small rotors, which currently only gives it a flying time of five minutes.

The drone's forward-looking infrared (FLIR) camera can find fires, estimate their temperature and help map their location on the ship. It can share that information with the SAFFiR humanoid, and could form a team to help sailors fight fires in the future.

"We're working toward human-robot teams" ONR program manager Thomas McKenna said in the release. "It's what we call the hybrid force: humans and robots working together."

SAFFiR has been under development for about five years. Researchers want to add natural language and gesture control as well as faster locomotion to make it an effective firefighting tool.

Tim Hornyak covers Japan and emerging technologies for The IDG News Service. Follow Tim on Twitter at @robotopia.

Join the Computerworld newsletter!

Error: Please check your email address.

Tags Carnegie Mellon UniversityOffice of Naval ResearchroboticsSensible MachinesVirginia Tech

More about IDGMellonNews

Show Comments