A NASA drone is built to divert itself when it judges its landing spot unsafe, a test for Moon and Mars landers

A four-rotor drone named Alta-X carried NASA’s landing-guidance payload over the dry lakebed at Edwards, California, and flew simulated lunar descents with it on Aug. 27. The payload maps the ground below, rates candidate touchdown zones and, when its chosen zone looks hazardous, steers the vehicle to a better one. NASA’s Armstrong Flight Research Center published its account of the work on Nov. 4.

Tech news worth reading: Get Techpresso, the daily email with the most interesting tech news and insights, read by 400,000+ professionals from Google, Apple and OpenAI.

Get Techpresso →

We may earn a commission if you sign up.

The aim is not to land a drone gracefully. It is to prove software and sensors that must someday set down a spacecraft where no pilot can intervene.

SPLICE sensors and software in flight

The package is called Safe and Precise Landing – Integrated Capabilities Evolution, shortened to SPLICE, and NASA’s Johnson Space Center in Houston developed it. According to the Armstrong release on the navigation experiment, the payload builds high-resolution digital elevation maps from lidar and other sensor data, assesses potential landing areas and then acts on the result. If the system determines the targeted landing area is unsafe, it diverts the drone to a safer location.

The behavior is described as a design capability. The release does not report that Alta-X actually aborted a landing and relocated during the flights, so the diversion remains what the system is built to do rather than an event NASA documented in this account.

The wider SPLICE project, described on NASA’s project page, combines several components: a Navigation Doppler Lidar that measures speed and altitude with laser beams, a Hazard Detection Lidar that scans the landing site in three dimensions for steep slopes and large rocks, a Descent and Landing Computer that fuses the data, and hazard-detection software that picks safe sites to divert to. Terrain Relative Navigation compares live camera images against orbital maps to fix position.

From a June bench check to a lakebed descent

Armstrong integrated, tested and flew SPLICE on the Alta-X in June and July 2026, checking communications, data handling, telemetry and command-and-control performance. The Aug. 27 flights then ran simulated lunar descent and landing maneuvers at Rogers Dry Lake, a flat expanse beside the center where NASA has tested aircraft for decades. A photo caption in the release dates hover-maneuver communication tests to July 15, which fits the earlier checkout phase, though the release text is not perfectly consistent about its calendar.

A larger rehearsal was planned next, a full-scale test at Voyager Technologies in Mojave, California. The point of working up in stages is cost. A small drone can repeat an approach dozens of times in a day, while a real lander gets one try.

Davis Adams, NASA’s project manager, called the flight “the culmination of almost a decade of work.” He said it raised the technology’s readiness level for future robotic and crewed lunar flights, and that the guidance and navigation validation applies to Earth, Moon and Mars landings.

The commercial hand-off and a decade of Flight Opportunities tests

NASA’s interest extends beyond one agency’s missions. According to Adams, NASA has worked with several commercial lunar lander providers to place iterations of SPLICE in their platforms, and some landers for upcoming Moon Base missions use data from systems like it. He added that the guidance system is planned for NASA’s software catalog this fall, and that some companies are evaluating its navigation for precision landing and autonomous surface mobility.

The program’s roots lie in earlier flights supported by the Flight Opportunities program, which tests promising technologies on suborbital and hosted orbital flights with commercial flight providers. The release notes a 2019 flight of a descent altitude and landing trajectory similar to a lunar mission and a 2020 demonstration of two NASA-developed precision landing sensor systems. NASA’s page on Flight Opportunities lists a hazard detection lidar system that has already traveled to the Moon among its successes.

The project page describes SPLICE’s goal as safer, more precise landings on the Moon, Mars and icy worlds, including hard-to-reach areas of scientific interest. It also names the possibility of landing sites near boulders and shadowed craters that Apollo crews had to avoid. Landing computers are part of that picture: a 2023 NASA technical paper on the Descent and Landing Computer describes a custom single-board computer built around a Xilinx UltraScale+ processor that runs heavy algorithms in real time and predictably, the kind of determinism a lander cannot do without.

Earlier validation used a suborbital rocket instead of a drone. A NASA and Blue Origin paper presented at the AIAA SciTech Forum reports that the SPLICE technologies flew as an integrated payload on Blue Origin’s New Shepard vehicle, with the authors gathering processor-utilization and bandwidth data in flight and in a hardware-in-the-loop simulation. The Alta-X campaign adds the piece the rocket flights could not offer, which is cheap repetition close to the ground.

The project page lists September 2026 as the closeout date for SPLICE as a development effort, which makes the planned software-catalog release this fall the next milestone to watch.

This article was produced with the assistance of AI and reviewed by Morning Overview editors prior to publication.

Tech news worth reading: Get Techpresso →

We may earn a commission if you sign up.