NASA's SPLICE drone test connects seeing the ground with choosing a landing

Photo: NASA/Ryan Kline, AFRC2026-0152-30. Supplied for factual editorial use under NASA media guidelines. No PING image edits. Image source · NASA use conditions. An Alta-X drone carrying the actual SPLICE experiment near NASA Armstrong in California on 15 July 2026. This is an Earth-based integration-flight photograph, not the 27 August descent test or a lunar landing. NASA supplied the full-frame 1920 by 1280 article version; the 7675 by 5117 original is retained separately.

Choosing where to land is a harder problem than reaching a marked point. A vehicle also needs to understand the ground beneath it and change course if that ground is unsuitable. NASA's latest SPLICE flight-research report makes those separate jobs visible: a drone in California is being used to explore technology intended for far more distant landings. The useful story is how sensing, computing and guidance are tested together.

NASA published its account on 4 November. It reports simulated lunar descent and landing manoeuvres on 27 August at Rogers Dry Lake, using an Alta-X drone with the SPLICE payload. The photograph supplied here comes from earlier integration flying on 15 July. These are different dates and stages, rather than a new November launch or evidence of a landing on another world. Read NASA's report.

A landing system is a chain of decisions

SPLICE combines sensors, a computer and algorithms. NASA's project page describes navigation Doppler lidar for estimating movement and height, hazard-detection lidar for building a three-dimensional view of the ground, and computing that evaluates possible landing locations. Its terrain-relative navigation compares live camera images with an existing map. Guidance then has to act on the result. An individual sensor reading is one part of this chain. How NASA describes the components.

That separation helps explain the experiment. Knowing where a vehicle is does not by itself establish that the intended landing point is clear. Detecting a hazard does not by itself show that the vehicle can reach an alternative. A useful demonstration must connect the observations to the decision and then to the vehicle's response. This is a way to read the engineering account, not a claim that PING has evaluated the control system.

What the Earth flight can establish

NASA describes terrain mapping, assessment of possible landing areas and diversion when a target is judged unsuitable. It also reports earlier integration checks between the payload and the aircraft. Those are the agency's account of the work; PING has not attended the flights, examined telemetry or reproduced the analysis. The report supports a particular development milestone, not a general guarantee of autonomous landing safety.

A recent report on a longer development programme

The project page dates SPLICE's introduction to October 2017. A separate NASA account from 20 June 2019 describes rocket-sled testing of navigation Doppler lidar: a stationary instrument measured a target moving along a track. That tested a different part of the problem from flying a payload on a drone. Keep those experiments distinct instead of treating every test photograph as interchangeable evidence. Historical mission dates in the 2019 article are historical plans, not current commitments. The earlier sensor-testing account.

The contrast matters when a development story travels beyond its original audience. A test can be useful because it isolates one difficult job, even when it does not reproduce the whole destination environment. Ask which hardware and software were connected, what the vehicle was asked to do, and what evidence would let a reviewer assess the result. Then keep the limits alongside the achievement.

Read the demonstration and the destination separately

For engineering students or small robotics teams in Accra, Nairobi or London, this is a concrete example of a system being developed through different kinds of experiments. These locations describe the readers, not a claimed local deployment or partnership. The transferable question is how a team moves from a component result to an integrated demonstration without losing track of what each test actually showed.

Before repeating a claim about an autonomous system, look for the test conditions, the basis for calling a result successful and the remaining work. A photograph can identify the equipment and setting; it cannot supply the missing measurements. In this case the image shows the real experiment's Earth-based aircraft. The assessment of its flight results remains attributed to NASA, with no PING accuracy figure or comparative ranking added.

SPLICE's long-term purpose is planetary landing. The current news is a step in testing the system that might support that purpose. Keeping the two scales separate makes the achievement easier to understand and the next milestone easier to question. For another way to explore actual mission records, see PING's Artemis II archive guide.