by

In Japan, Rescue Robots Are Poised to Go From Lab to Quake Scene

Last Thursday night, Robin R. Murphy, director of the Center for Robot-Assisted Search and Rescue at Texas A&M University, held a goodbye party in College Station for Japanese robotics researchers who had come to the center for workshops on using their creations in an emergency.

The next day, the workshops became reality. The massive earthquake and tsunami that ravaged Japan that Friday meant the scientists, already booked on a plane, were rushing home to help.

Satoshi Tadokoro of Tohoku University, based in Sendai, one of the cities hardest hit by the disaster, raced back with his team to offer the use of a robot developed at Tohoku. The snakelike robot can enter tight spaces and use a camera to survey them, something particularly helpful in collapsed buildings, says Ms. Murphy.

Another team, led by Eiji Koyanagi, deputy director at the Chiba Institute of Technology’s Future Robotics Technology Center in Tokyo, offered use of his Quince robot. It’s a small, tanklike machine with treads that can maneuver through rubble, and it has sensors that can detect and report chemical contamination.

These and other robots will probably not be used in the initial days, but they could be called into action soon, as teams begin to search for victims in less-accessible places and assess the damage to collapsed buildings and flooded waterways, Ms. Murphy says.

The most useful robots at the moment, she says, will be aerial drones that can provide instantaneous assessments of the damage to upper levels of buildings, and remotely operated underwater vehicles, or ROV’s, which can help locate underwater objects and determine the condition of bridges and pipelines.

“We want these to be tools that extend ourselves in real-time,” she says.

The aerial robots, which can provide live feeds, are an improvement over helicopters and other manned vehicles, which deliver results when they return to ground. The robots can also hover in place, rather than circling around, which is better for monitoring a location over time.

The underwater robots function as an “acoustic camera,” using a sonarlike technology to provide black and white images of conditions underwater. The alternative is to send divers into waters that often have extremely strong currents and can be hard to see through because of all the debris.

Ms. Murphy, a professor of computer science and engineering at Texas A&M, says she was drawn to developing robots for rescue situations after the 1995 bombing of a federal building in Oklahoma City. One of her graduate students went to assist in the recovery and reported back that robots could have helped search for survivors.

She says there’s still resistance among rescue teams to using robots in early stages of disasters. That’s because humans and dogs cover ground more quickly.

But in situations where getting into small spaces are crucial, Ms. Murphy says, the machines could be used earlier than they currently are. “One of the depressing things in the work that we do is that robots are called as the sort of Hail Mary pass,” she says.

She hopes the use of robots will grow to be more accepted, but she’s also working to make rescue robots more accepted by the people who rely on them. As part of a National Science Foundation grant, she and Clifford I. Nass, a professor in the communications department at Stanford University, are working to make rescue robots appear less “creepy,” so that the people they rescue respond more positively to them.

They hope to create robots that can modulate the volume of their voices, light up for better visibility, make better eye contact and even pipe in multimedia, such as video or music, to make long rescue efforts more bearable.

This is important because research she and Mr. Nass have conducted indicates that people attribute human characteristics to robots and often describe them as “creepy,” even though they know the robots are not animate creatures.

If victims become frustrated or uncomfortable with a robot trying to help them, it can make them resistant to accepting the robot’s help, she says, and that can be disastrous in an emergency.

This research won’t be able to help in Japan—the current generation of robots don’t have speakers, for instance. Now, Ms. Murphy hopes to bring over a team using more updated technology to help.

“We fully expect to go out there,” she says. “During the recovery phase, there’s a lot to be done to assist our Japanese colleagues.”

Return to Top