Robots and drones are traditionally deployed to do jobs that humans find dirty, dull, or dangerous. The Office of Naval Research (ONR) has revealed a robot that fits that last job description perfectly: a walking, humanoid robot prototype that uses thermal imaging to identify overheated equipment and wields a hose to quench fires.
The catchily titled Shipboard Autonomous Firefighting Robot (SAFFiR) is a humanoid robot developed by researchers at Virginia Tech. The 65 kg SAFFiR comes equipped with infrared vision, which allows it to to see through dense smoke, and a rotating laser for light detection and ranging (LIDAR). Google’s self-driving cars also use LIDAR for navigation.
“We set out to build and demonstrate a humanoid capable of mobility aboard a ship, manipulating doors and fire hoses,” said Thomas McKenna, ONR program manager for human-robot interaction and cognitive neuroscience, who recently helped demonstrate SAFFiR’s capabilities, putting out a small fire aboard the USS Shadwell, a decommissioned Navy vessel.
Navigating around a complex environment is quite a feat for any robot, especially one that may be fighting fires in a ship rolling on a pitching sea.
“Balancing on any type of terrain that’s unstable—especially for bipedal robots—is very difficult,” Brian Lattimer, associate professor for mechanical engineering at Virginia Tech, said in a statement. “Whole-body momentum control allows for the robot to optimize the locations of all of its joints so that it maintains its center of mass on uncertain and unstable surfaces.”
Like Atlas, a similar-looking rescue-focused robot designed for the U.S. military, SAFFiR was first demonstrated in DARPA’s 2012 Robotics Challenge, which was motivated by the need to navigate radioactive hotspots, like the damaged Fukushima nuclear reactor. The broader idea is to outfit robots like SAFFiR with longer-lasting batteries, advanced intelligence, and sensors so they can detect and fix other technical problems in challenging environments, lessening the load on soldiers and sailors.
With the military developing fighting robots, Uber partnering with CMU to develop driverless car technology, and even plant nurseries deploying fleets of pint-sized robot farmers, is the writing on the wall for the most physical jobs?
“Don’t think it’s going to replace you: It’s going to assist you,” said John Seminatore, a graduate assistant at Virginia Tech. “Every sailor is trained as a firefighter, but that’s not their expertise, that’s not what they’re doing every day.” He points to the support that the Navy provided in the aftermath of the 2011 Haiti earthquake: “You’ll be able to send [SAFFiR] into dangerous situations without having to send a sailor in there first.”
In fact, it may be the jobs that don’t involve any hardware or human power at all that are at most risk of being automated. “The next frontier for automation is non-routine work,” Matt Beane, a researcher at MIT’s Sloan School of Management, told me in 2012. “Some of the biggest changes in work could be at the high end. These jobs can be automated without a physical avatar. Once you have got good AI, it’s replicable at almost zero cost.”
So while taxi drivers, farm workers, and firefighters can be replaced by robots operated by artificial intelligence (AI) or a human operator, it’s still expensive to manufacture that hardware. Replacing lawyers or software developers or journalists, however, could result in the ultimate scalable business.