It is likely the most sophisticated half-liter scoop ever built -- the bucket that will dig up dirt from the martian polar surface and dump it into the spacecraft's soil analyzer.

The cup that is mounted at the end of the lander's robotic arm is equipped with special alloy blades engineered to dig through some of the hardest frozen soil scientists could imagine. Nobody knows just what kind of surface to expect at the south polar landing site, so engineers have had to prepare for anything.

For about 8 months during the 4-year design, construction and testing of the robotic arm, engineers even employed a full-time chemist to make the kinds of soil that the Mars Polar Lander might encounter on the Red Planet.

"Her only focus was to become a soil-simulant expert," said Jeff Slostad, a member of the robotic arm team who worked on the arm's digging bucket. "She became basically a cook, mixing fine particles such as clay, and very fine particles of silica sand, mixed with things like baking soda and salt."

The baker was Jennifer Herman, a chemist at the Jet Propulsion Laboratory who learned almost everything scientists know about martian soils in order to make a variety of soils that the arm team could use to test their shovel.

Herman started out with sand that had a particle size similar to what scientists think exists on Mars, then added clays that reflect light the same way soils on Mars appear to do. She added salts to make the soils clump more, and baking soda to make samples fluffy.

"They don't know what the compressive strength of that soil is going to be," Herman said, "They don't know if the lander will just land on a very firm surface or if it is going to sink. And that's what a lot of people said about the moon too: 'Is it going to be 50 feet of fluff, or 5 millimeters or what?' So we tested fluffy stuff and we tested very very hard stuff."

All the samples were then frozen to minus 85 degrees Celsius (minus 121 degrees Fahrenheit) -- the coldest temperatures expected on Mars. In this chill, Slostad and members of the robotic arm team slowly perfected their designs.

The robotic arm is a little more than 2 meters long and is designed so that it can dig a trench as deep as two feet (about half a meter) into the surface -- if the soil is soft enough. It has about the same digging strength as a human arm. The bucket will scoop up samples of the martian dirt and dump them into chambers in the lander's Thermal and Evolved Gas Analyzer (TEGA) -- an instrument that will cook soil samples to test their composition. (Click here to see a demonstration of the robotic arm's bucket dumping.)

"If the surface we land on is just this hard, concrete-like surface, it's going to take us a while just to get a sample," Slostad said. "Our challenge then isn't to get deep into the surface, it's to grab a sample off of a very hard surface with the same arm that's designed to work in the very loose stuff."

"Usually if someone's persistent -- and a robot usually is pretty persistent -- you have a good chance to get through pretty much anything that you come across, particularly if you have nice tools. And so we put out some really nice tools on the bucket so it can dig into whatever it comes across," Slostad said.

To get through any unyielding surface, the robotic arm team built its cutting tools out of tungsten carbide, the same metal alloy used in tools that cut steel. These were designed so that they would dull very little with use, a feature developed under the guidance of engineers at several leading knife manufacturers, Slostad said.

One of the cutting tools resembles the face of a cheese grater, and the edge of the bucket has prongs called ripper tines that are similar to the gouges found on the back of backhoe buckets.

"You know the little three-prong hand tool that you use in the garden? Same sort of theory," Slostad said. "Put all of your force in one tiny little area."

Digging is only one of the many tasks the robotic arm will need to accomplish. It needs to dump accurately into the TEGA sample chambers, a process that has been tested just as exhaustively as the digging. (Click here to see a photo animation of the dumping bucket at work)

Team members spent a month testing dirt dumping at low temperatures, and even conducted dumping experiments in simulated martian gravity. Members of the robotic arm team flew aboard NASA's KC 135-A -- an airplane that uses steep descents to produce the effects of low gravity -- and practiced dumping various fine-grain materials such as talcum powder to learn about soil mechanics in low gravity.

Slostad said he and his team are confident the robotic arm will do a good job finding, digging and delivering plenty of sample soil to the TEGA, no matter what the surface turns out to be. Others agree.

Hop Bailey, the program manager for the TEGA instrument called the robotic arm elegant and effective.

"When you look at the design it's kind of surprising that it works as well as it does," Bailey said. "There are no grinders or mechanical devices that are scraping and stuff like that. It's just the design of the front of the scoop has tines that give you a combination of penetration and scraping. They get a lot of material. When they dump on us we get submerged by all this stuff they dump."