Several years ago the Department of Homeland Security approached the National Institute of Standards and Technology with a challenge. DHS wanted to adopt new technologies to help emergency responders do their job more efficiently while keeping them out of harm's way.
Robots were the obvious answer. But no one was quite sure how well they might work in emergency response scenarios, said Elena R. Messina, acting chief of NIST's Intelligent Systems Division.
"The technology was not proven," Messina said. "They work in an environment that is always changing with many different alternatives. DHS was looking for a way to evaluate the robots before they made any purchasing decisions."
DHS wanted to be sure the rescue responders were getting products that were dependable. The result is a series of robot evaluation exercises for urban search and rescue teams, the most recent held in November at Texas Engineering Extension Services' Disaster City training facility in College Station, TX.
Since the Disaster City testing several of the test methods have been approved by the ASTM (American Society for Testing and Materials) International Committee on Homeland Security, Operational Equipment. ASTM International is one of the largest voluntary standards development organizations.
?"We've had a couple approved and several of the ones at Disaster City are going to be balloted soon," Messina said.
Forget "Star Wars." The response robot evaluation exercise at Disaster City involved 35 of the technologically advanced robots that exist in the real world, complete with antennas, manipulators and sensors. On hand were ground based portable robots made to navigate large unknown situations, highly agile robots large enough to remove a human from complex environments and confined space accessible robots that can be deployed into small voids or tossed into inaccessible areas.
Add to the assortment aquatic vehicles with sonar and other sensors and quad-rotor aerial vehicles. In all five categories, the robots made use of color cameras, two-way audio, thermal imagers, chemical sensors, 3D mapping and GPS/GIS location.
Working with USAR teams to develop performance standard, DHS and NIST will, in turn, develop comprehensive standards related to the testing and certification of effective robotic technologies. These standards will address robot mobility, sensing, navigation, planning, integration into operational caches and human factors.
Such standards will allow DHS to provide guidance to local, state and federal homeland security entities regarding the purchase and deployment of robotic systems for US&R applications.
"We are working with the responders," Messina said. "They let us know who they think could use the robots. We try to translate their needs into test methods. You have to go from this very broad domain of all the things you think they can do to specific performance standards."
One practice scenario featured ground robots working in confined spaces within partially collapsed structures. Another involved down-range reconnaissance of train wrecks involving hazardous materials with the robot operator more than 500 feet away, said Adam Jacoff, an NIST robotics research engineer.
"Responders are not allowed to go down range without suiting up in protective gear," he said. "So the idea is to get quick reconnaissance of what we are dealing with while responders are getting dressed. You send a robot down to check out the placards and use all types of sensors."
The basic requirement is four gas sensors. More than that is considered value added, Jacoff said.
"These sensors need to be integrated into the robot systems so the displays are clear and understandable," Jacoff said. "The radio command has to handle robots going down range as much as 1,000 feet from the operator."
These and other tests capture statistically significant performance data on the robots, Jacoff said.
"The performance data will inform the buying decisions," Jacoff said. "At the same time the test methods will give the developers a target to design towards."
Different robots present different issues, Jacoff said. Many of the robotic systems operating at the evaluation exercise use radio command in which the operator interfaces with the machine from a remote location.
"In a refinery setting this presents a problem," Jacoff said. "There is a lot of metal that makes it a real problem to work in that kind of environment."
Two separate tests are used for radio command devices. First, a line of sight test is done. To complicate matters, a wall about 24-feet high and 80-feet wide is built using shipping containers.
"The operator works behind it and we see how the radio attenuates on the other side," Jacoff said. "That is important for your refinery-type applications because these metal environments are radio difficult. We have all the best military radios here going through the same test method."
Countries represented at the November exercise included the U.S., Japan, Germany, China, Canada and Switzerland.
Disaster City is a 52-acre training facility designed to deliver the full array of skills and techniques needed by urban search and rescue professionals, including full-size collapsible structures that replicate community infrastructure, including a strip mall, office building, industrial complex, assembly hall/theater, single family dwelling, train derailment and three rubble piles.
Testing at Disaster City offers scientists a wide variety of situations, Messina said.
"If they want to clear a building they can send in the robot to map it," she said. "It could create a map so the responders would get a reasonable representation of what is in the building and plan how they would try to evacuate people."