OK, so no one has gotten to say it yet, but they might get to in the near future. A new breed of robot rescuers is being tested in order to take some of the strain and risk off of the shoulders of human first responders, and get to trapped people in places where humans would never be able to go, or find potentially dangerous chemicals.
The National Institute of Standards and Technology (NIST) held a rescue robot exercise in Texas last week in which about three dozen robots were tested by developers and first responders in order to develop a standard suite of performance tests to help evaluate candidate mechanical rescuers. This exercise was sponsored by the Department of Homeland Security’s Science and Technology Directorate to develop performance standards for robots for use in urban search and rescue missions.
This also represents a new level of equality in robotics research, where data can be compares in an apples to apples and not an apples to oranges situation. “It is challenging to develop the test standards as the robots are still evolving,” explained Elena Messina, acting chief of the Intelligent Systems Division, “because standards are usually set for products already in use. But it is critical for developers to be able to compare results, which is not possible without reproducible test environments. So, we have reproducible rough terrain that everyone can build in their labs, whereas you can’t reproduce a rubble pile. This way, developers in Japan can run tests, and people in Chicago can understand what the robot achieved.”
Exercises included testing battery capacity by having robots perform figure eights on an undulating terrain and mobility tests in which robots ran through increasingly challenging exercises beginning with climbing steps and escalating to climbing ramps and then making it up steps with unequal gaps. A new mapping challenge introduced at this event tests how accurate a robot-generated map can be—the robot must traverse a simulated “wooded area” that has uneven terrain and PVC pipes for trees, and create a map using its sensors.
If you happen to be wondering what that looks like you can see the video at the Disaster City TEEX Web site: www.teexblog.blogspot.com/. I strongly recomend checking it out.