NASA had planned to launch its first Mars sample-return mission in 2005. But back-to-back losses of the Mars Climate Orbiter and Mars Polar Lander in 1999 prompted the agency to pause and rethink its plans for exploring the Red Planet.

Even with an extra six years to get ready for the first sample-gathering mission, NASA officials say they have no time to waste.

"The set of technology needed for sample return is probably the most challenging one there is," said Scott Hubbard, Mars Exploration Program director at NASA.

Of its roughly $400 million annual budget for Mars exploration, NASA intends to devote about 10 percent to technology development, Hubbard said.

"Its our goal to have a robust investment," Hubbard said.

While some of the money will be doled out to NASA field centers to fund in-house technology development efforts, NASA also plans to issue a number of competitive solicitations to industry, he said.

Hubbard said some of the technologies NASA needs for the sample-return missions would get their first test in 2007, when the agency plans to launch a long-range science rover to Mars. The most challenging technologies are those needed for precision landing, he said.

Hubbard said the 2007 lander must be capable of touching down within 3.1 miles (5 kilometers) of its designated target. NASAs successful Mars Pathfinder, by contrast, was accurate to within plus-or-minus 60 miles (100 kilometers).

The lander also must be equipped with a hazard avoidance system so that it doesnt attempt to touch down on a steep slope, large boulder or any other mission-ending obstacle, Hubbard said.

The list of technologies needed for the 2011 sample-return mission goes well beyond what is needed for the 2007 rover mission, Hubbard said. NASA must find a way to select and collect worthwhile samples, prevent biological contamination, then launch from the Martian surface a sample-return canister that must dock with an orbiting mother ship, which will return to Earth, he said.

The most daunting of those challenges, according to Hubbard, is the Mars orbit rendezvous between the sample canister and the mother ship.

NASA is evaluating a proposal submitted last year by the agencys Johnson Space Center in Houston to do a direct-to-Earth sample return, thereby eliminating the need for an orbital rendezvous. Nevertheless, the agency is moving ahead under the assumption that the samples will be launched from the surface in a so-called Mars ascent vehicle and captured by an orbiting spacecraft and brought back to Earth.

Hubbard said the mini-launchers likely will bring their fuel from home rather than making it on the surface from the Martian atmosphere. Some technologists see so-called in situ propellant generation as a key to reducing spacecraft weight, but Hubbard does not think the technology will be ready by 2011.

"Based on what we know today, I think it's probably not the first choice we would make," he said.

But NASA would like to make use of in situ resources soon. Doug Cooke, manager of the exploration office at Johnson, said the Mars ascent vehicle as envisioned would be compatible with the kinds of fuels that can be generated from Mars atmosphere. When in situ propellant generation is ready for use on Mars, the mini-launchers will be ready to take advantage, he said.