Lunar Precision Landing: NASA Tests SPLICE Hazard Avoidance on a Drone
NASA's November 4 SPLICE report adds an Earth-based hazard-avoidance flight test to the lunar precision-landing story, with software availability and flight qua
Updated November 7, 2026 NASA's November 4 report adds a real flight demonstration to this explainer: an Alta-X drone flew simulated lunar descent and landing maneuvers with the SPLICE guidance, navigation, and control payload. It was an Earth-based test, not a Moon landing or a certification of a commercial lander. SPLICE moves from sensing hazards to steering around them NASA says the August 27 flight at Rogers Dry Lake near Armstrong Flight Research Center in California used SPLICE to map terrain and assess possible landing areas. The system is designed to divert the vehicle when its targeted landing area is unsafe. Its digital elevation maps turn sensor observations into a practical choice about where to touch down. The result gives the infrastructure argument below a concrete test case: identifying a rock or slope is useful only if guidance can turn that information into a safe, achievable trajectory. NASA's November 4 flight report describes the demonstration. Johnson Space Center developed the payload and worked with Armstrong specialists to integrate it on the drone. NASA reports June and July tests of communications, data handling, telemetry, and command-control performance before the descent demonstration. Those interfaces matter: a navigation solution has to reach the flight-control system in time, in the expected format, with enough reliability to guide the vehicle. A good terrain map cannot compensate for a broken command path. What each part contributes NASA's SPLICE technical overview separates several functions that are often bundled into the phrase autonomous landing. Navigation Doppler Lidar estimates altitude and velocity from reflected laser light. Hazard Detection Lidar constructs a three-dimensional terrain map. Hazard Detection Software uses that map to identify unsafe regions and candidate safe sites. The Descent and Landing Computer processes sensor information and runs the algorithms that support the landing decision. Terrain Relative Navigation compares camera observations with stored orbital imagery. Active Terrain Relative Navigation instead extracts three-dimensional features from elevation maps to estimate location and build a more continuous terrain model. Dual Quaternion Guidance calculates a fuel-efficient trajectory subject to vehicle and environmental constraints, keeps the sensors pointed appropriately, and can redirect the vehicle toward a newly selected site. This is why precision landing is a coordinated system rather than a single better camera. What the demonstration does not establish An Alta-X flight in Earth's atmosphere cannot reproduce lunar gravity, vacuum, rocket-plume interaction with regolith, or the full range of polar illumination. The November report says the flight raised technology readiness, but does not give a numerical readiness level, measured touchdown accuracy, success rate across a test population, or a complete operating envelope. Those missing metrics should not be replaced with invented precision figures. The demonstration supports further integration; it does not certify that any named lander can now deliver cargo beside a habitat. The two NASA pages also need careful handling as a timeline. The November 4 article refers to a full-scale Voyager Technologies test as occurring "later this month." The technical overview separately lists August 2026 closed-loop drone testing at Voyager's Lunar Surface Proving Grounds in Mojave, involving the guidance and active navigation systems, and lists SPLICE project closeout in September. These are different published descriptions with an unresolved timing inconsistency. This update does not treat the relative-date wording as proof of a new completed November campaign, or merge the two sites into one flight. The next step is usable software and vehicle-specific evidence Project manager Davis Adams said NASA was coordinating with industry on use of the guidance system once it becomes available in the NASA software catalog, with availability planned for the fall. The same report says some companies are evaluating navigation capabilities for precision landing and possible autonomous surface mobility. A planned catalog release is not confirmation that the software is already downloadable, licensed, integrated, or qualified on a flight vehicle. The article does not identify all prospective users or promise a deployment date. For lunar logistics, the useful follow-up evidence will be measured landing error, hazard-detection performance, processing latency, fuel cost of diverting, and behavior when a sensor or map becomes unreliable. Operators also need the allowable lighting, terrain, and approach conditions. Those requirements connect the flight test to the original point of this explainer: repeat access needs maps, onboard decisions, communications, and site rules that work together. A successful drone demonstration advances that chain without replacing the remaining flight and operational qualification work. Sources checked November 7, 2026: NASA Armstrong's November 4 flight report and NASA's SPLICE technical overview. Test dates, future availability, and unresolved source chronology are distinguished above. Existing illustration retained. A lunar landing is not finished when a vehicle reaches the Moon. The hardest part is often the last few kilometers, when the spacecraft has to recognize terrain, avoid hazards, manage dust, and touch down close enough to useful infrastructure without relying on GPS. That is why precision landing and navigation are becoming core cislunar infrastructure. South-pole missions need more than courage and propellant. They need sensors, maps, timing, communications, beacons, and agreed operating rules that make repeat landings safer and more predictable. AI-generated image Precision landing turns site access from a one-off spacecraft maneuver into reusable lunar infrastructure. Why the final descent is different at the Moon Earth aircraft and drones operate inside a dense navigation environment: GNSS, surveyed runways, air-traffic procedures, weather networks, radar, radio aids, and ground crews. A lunar lander has none of that by default. It arrives at a world where lighting can be harsh, shadows can hide hazards, dust can obscure the surface, and communications may be blocked by local terrain. The lunar south pole is especially demanding. It is attractive because some ridges see long periods of sunlight and some nearby permanently shadowed regions may preserve water ice. The same geometry creates low sun angles, deep shadows, and rugged terrain. A landing zone that looks acceptable from orbit may still contain boulders, slopes, craters, or dust conditions that matter during the last seconds of descent. Precision also matters because future missions will not be isolated. A cargo lander may need to arrive within rover range of a crewed outpost. A power unit may need to land close enough to a cable route. A communications relay, science payload, or resource prospecting package may have to hit a narrow operational box. Missing by kilometers can turn a successful landing into a poor infrastructure outcome. Map Orbital terrain and lighting products Sense Lidar, cameras, radar, and altimeters Match Terrain-relative navigation Avoid Hazard detection and divert logic The basic stack The first layer is mapping. High-resolution orbital imagery, elevation models, illumination models, and thermal data help mission teams select candidate sites and build onboard references. Better maps reduce uncertainty, but they do not remove the need for onboard judgment. Lighting changes, dust, and local terrain can still force a vehicle to make decisions in real time. The second layer is relative navigation. A lander can compare camera or lidar observations with stored terrain maps to estimate where it is during descent. NASA's terrain-relative navigation work for Mars showed the value of matching sensed terrain to onboard maps. Lunar systems ada