Landing is the toughest part of any trip to Mars. The planet's atmosphere is too thin for parachutes alone to bleed off a spacecraft's blistering entry speeds, and landing solely via retro-rockets requires more fuel than any near-future mission to Mars is likely to have. NASA uses both techniques together for most of its Mars missions, and even its high-tech Curiosity rover used a vintage 1970s parachute apparatus during its landing in 2012. That technology limits landings to 1.5-ton probes. Anything heavier will yield only a smoking crater. For more robust robotic missions—or eventual human ones—a new approach is required.

This month the space agency plans to test the full conception of a new lander, the Low-Density Supersonic Decelerator (LDSD), which could deposit twice as much mass across a wide variety of Martian terrain. The LDSD consists of a six-meter Kevlar-insulated inflatable disk and a 30-meter parachute capable of withstanding supersonic speeds—the largest ever deployed. An eight-meter disk suitable for human landings is also under development. Together the relatively lightweight inflatable disk and giant parachute could deliver bulkier payloads to Mars, without requiring large amounts of extra fuel or very heavy atmospheric-entry heat shields.

To test the LDSD, NASA will use Earth's upper stratosphere as a stand-in for the thin Martian atmosphere. First, a giant, football-stadium-sized balloon will hoist a mock lander 37 kilometers high over the Pacific Ocean near Hawaii. Then the LDSD will separate from the balloon and fire rockets to reach higher altitudes and speeds. At a height of 55 kilometers and a velocity nearly four times the speed of sound, its disk will inflate, slowing the vehicle so that the parachute can deploy. This will be the critical moment—aerodynamic forces shredded the parachute during an early flight demonstration in 2014.

If all goes well this time around, about three hours after launch the saucer will splash down in the Pacific, clearing the way for bigger, more sophisticated missions to Mars and other destinations in the solar system. And even if there are hiccups, says Ian Clark, the LDSD lead at the NASA Jet Propulsion Laboratory, “we would much rather learn these items now than learn them at Mars.”