The situation after an earthquake is chaotic, with people trapped, buried under rubble. Even without the aftershocks, more buildings are in danger of collapsing. It’s not an easy situation for rescue teams; they dig for buried victims and tend to the injured but are always in danger of being buried alive themselves. At this point it would be wonderful if robots could take over the most dangerous work. Which they can. Such high-tech devices are already being used—and they’re getting smarter.

The first use of such rescue robots dates back eighteen years. After the terrorist attack on September 11, 2001, they were scrabbling through the rubble in large-scale operations. Dr. Robin Murphy and her team from the Center for Robot-Assisted Search and Rescue (CRASAR) were also there. They guided the mechanical assistants through the ruins and transmitted live images from the robots to the rescue teams, providing the precise information necessary for making correct decisions from a safe distance.

Murphy has been researching robots for more than twenty years. She is one of the pioneers in this field. The first time she starting thinking about rescuers made of plastic and steel was in 1995. A truck bomb terrorist attack on the Murrah Federal Building in Oklahoma City, Oklahoma, had shaken the United States. One of Murphy’s students helped with the rescue there and told Murphy about how helpful it would have been to have had small, maneuverable robots to perform certain tasks.

This gave Murphy the initial inspiration. Together with Satoshi Tadokoro, she got the ball rolling in a new field of research into rescue robots in the years that followed. Since then she’s helped with more than 25 rescue operations all around the world. “The robots themselves aren’t anything special,” she says. “But using them is.” Since the surroundings have usually been completely destroyed by a catastrophe, the experts must redefine the robot’s tasks on location for every new mission. Murphy now has three types of robot in her repertoire: helpers in the air, in the water and on the ground.

The most important component for a robot is its software: the artificial intelligence. Because hundreds of hours of video material are easily recorded during disaster operations, intelligent and learning algorithms search through this data and send it to the rescue teams pre-sorted. This allows the teams to more quickly evaluate the images and make better decisions.

The robots have yet to be able to work without human assistance. They are remotely controlled and deliver their sensors’ data to the rescue teams. The situations during catastrophes are simply too chaotic for robots to operate autonomously. The robots are also quite small due to the special circumstances of their use. “They have to fit inside the back of an all-terrain vehicle,” Murphy explains. “Everything has often been destroyed in disaster areas. Transporting anything large and heavy isn’t practical.”

Murphy has ambitious goals for the continued development of her robot rescuers. She plans to utilize the enormous progress made in sensors, microprocessors and artificial intelligence to support rescue teams even more effectively. It’s about getting the right information to the right people at the right time so they can make the right decisions. “My goal is for the robots to be routinely used in rescue missions by 2030,” she says.