As Senate hearings get underway this week to probe the accident at the Crandall Canyon mine in Utah that claimed the lives of six miners and three rescuers, attempts are also being made to evaluate the performance of robotic equipment sent in to assist the failed rescue mission.

The robotics team faced several daunting engineering and device-control challenges as it tried to unearth clues as to the whereabouts of the miners trapped 1,500 feet underground. The mine's depth meant that any mobile robot deployed needed to be equipped to travel not only through one of the seven boreholes drilled into the mountain above the mine but also to navigate for several hundred feet once it reached the mine's floor. In addition, the robot had to be waterproofed so that it would not malfunction when submerged in the pools of ground water it would encounter while traversing the mine's interior. And all of this had to be done using equipment above ground that would control the robot's movement and onboard cameras.

Workers, handicapped by time constraints and the continued shifting of the mountain's mass, were able to get only one mobile robot through a borehole and onto the mine's floor, where it traveled as far as seven feet from the point of entry, says Robin Murphy, director of the Institute for Safety Security Rescue Technology at the University of South Florida. The U.S. Department of Labor's Mine Safety and Health Administration, charged with overseeing all rescue and recovery operations in the aftermath of the August 6 cave-in, asked the Institute's Center for Robot Assisted Search and Rescue for help shortly after the accident.

The Crandall Canyon mine, located 120 miles south of Salt Lake City, crumbled in a cave-in so powerful that it registered a magnitude of 3.9 on the Richter scale. Tremors continued for several days after the initial cave-in, the strongest striking on August 16 with a magnitude of 1.6 and causing a subsequent collapse that claimed the lives of three rescuers. After that, rescuers relied primarily on the boreholes drilled throughout the mountain to give them some idea of where the miners might be trapped. Each of the seven boreholes, which ranged in size from 2.5 to nearly nine inches in diameter, took at least 40 hours to drill. The deepest hole cut more than 2,000 feet into the mountain and took more than 58 hours to drill.

Rescue crews decided to build a robot crawler using four-by-15-inch tracks from Inuktun Services that let the vehicle squeeze through tight spaces. The robot was attached to a control cable supplied by PipeEye International and outfitted with an Inuktun Spectrum 90 amphibious pan, tilt and zoom camera as well as a very small video camera.

PipeEye was on-site at the mine for seven days working with Inuktun's technology tethered to more than a mile of fiber-optic and copper cabling. While none of PipeEye's robotic devices were small enough to fit down the boreholes, the company was able to transfer the control electronics from one of its robots to one of Inuktun's devices.

Search teams experienced quickly deteriorating conditions as the mountain shifted and settled atop the mine. Although seven boreholes were drilled, the sixth and seventh collapsed before the robotic device could traverse them.

PipeEye's robot managed to touch down on the mine's floor through the fourth borehole. Its cameras captured images of wire mesh that had been used to stabilize the mine's ceiling but were instead hanging down to the floor. "All we could see was piles of rocks and mud," says James Milward, PipeEye's senior technician. Conditions on the mine's floor "were worse than we expected," Murphy adds. With continued seismic activity blowing holes in the mine's walls, the floor was filled with water and debris.

Unable to provide the rescuers with much information, the robot was left down the borehole overnight. Unfortunately, when the crew returned to retrieve it the next morning they discovered that the hole had shifted and the robot could no longer make its way back to the surface. Despite several attempts to remove the robot, including using a 400-pound chisel to break up the ground blocking its escape, the $35,000 robot became permanently trapped 52 feet below the surface. "It's in the mountain forever," Milward says.

Given more time, Milward is convinced he and his team could have designed a robotic device better suited for the situation. The robot used was just barely small enough to fit through the borehole. "If we do anymore work in this industry,'' he says, "we would use a smaller version."

Murphy hopes that lessons learned at Crandall Canyon will be incorporated into any standards that the U.S. Commerce Department's National Institute of Standards and Technology, or NIST, develops for future rescue robotics. Since every disaster is different, the best robotics designs give rescuers the most flexibility, she says, adding, "You never get it all right, even if you think you know what's down there."

It is crucial that robotic rescue devices be well crafted before they are sent into harsh conditions such as those found at Crandall Canyon. This includes having a watertight body with sensors and camera equipment positioned higher up on the robot so that they are less likely to be blocked by debris. All of this needs to be done to help workers on the surface operating the robot. The long hours and harsh conditions at a rescue site make it "unreasonable to expect that people will be able to compensate for any of the technology's shortcomings on-site," Murphy says.