Lunar PlanetVac

NASA/Firefly Aerospace. Public domain (NASA / US government work).

Overview

Lunar PlanetVac is a pneumatic sample acquisition and delivery system built by Honeybee Robotics and flown on Firefly Aerospace’s Blue Ghost Mission 1 to Mare Crisium under NASA’s Lunar Surface Instrument and Technology Payloads program [9]. It has no scoop, no arm and no moving sampling element. A jet of gas inside a sampling head lofts regolith into a transfer tube, and a capture system at the far end separates the sample from the flow.

The system is three parts: the sampling head, which needs access to the surface; the transfer tubes; and the capture system, which delivers the sample to an instrument or a return container [4]. The head can be built into one or more lander footpads or deployed on a boom. On Blue Ghost it rode on a Firefly-provided Sample Access Arm released by launch lock after touchdown, with the transfer tube running up the arm to a collection system on the temperature-controlled mid-deck. It flew alongside nine other payloads, among them the LISTER heat flow probe, on the same lander and to the same landing site [8].

Acquisition took place on 3 March 2025, the day after landing [13].

Specifications

Parameter	Value	Source
Subsystems	sampling head, transfer tubes, capture system	[12]
Working fluid	gas from a standalone canister, nitrogen on this flight	[13], [4]
Acquisition time	about 5 seconds from gas release to captured sample	[12]
Transport time	under 1 second from head to collection container
Sample verification	infrared break-beam pair at the chamber inlet, plus imaging
Capture mechanism	etched screens along a labyrinth path
Particle sorting	second gas burst, separation about 1 mm
Deployment on Blue Ghost	Firefly Sample Access Arm, launch-lock released

Mass, gas quantity, gas pressure, power draw and the dimensions of any part of the system are not published.

Mission profile

Parameter	Value	Source
Lander	Firefly Aerospace Blue Ghost	[9]
Launch	15 January 2025
Landing	2 March 2025, Mare Crisium
Sampling	3 March 2025
Volume collected	about 6.5 cc against a 1 cc requirement	[13]
Mass collected	13 g including 5 g of rocklets about 2 mm across	[12]
Sampling head position	not flush with the surface	[13]

The volume and the mass come from two abstracts by the same team and are not reconciled in either; neither publishes the bulk density that would connect them, so they are two independent statements about one event rather than a consistent pair.

Approach

Gas jets inside the sampling head point downward, loft regolith and sweep it into the transfer tube [4]. The efficiency argument is that gas behaves like an explosive in vacuum, so a small quantity moves a large amount of material, and the process is indifferent to gravity. Collection finishes in seconds after a single valve command, so the power draw and the operational overhead are both small and no operator is needed in the loop. OSIRIS-REx’s TAGSAM sampler works on the same principle at a larger scale, firing nitrogen through a brimmed head on contact with an asteroid rather than resting on the surface to sample; its qualification campaign found collection rising by a factor of four to five under reduced acceleration compared with the same test at one gravity, which is the general case for gas-driven sampling that Lunar PlanetVac’s own lunar test data does not yet cover [1].

The architectural claim is about routing rather than about digging. A scoop on a robotic arm can deliver only where the arm’s kinematics reach, and the instrument must be placed accordingly; a pneumatic transfer line can be run around obstacles, so acquisition hardware goes where sampling is best and instruments go where analysis is best [4]. Most excavation systems proposed for lunar in-situ resource work instead couple a cutting or digging element directly to a mobility platform, and a 2020 review of thirteen such systems found that none of them reported a consistent set of performance parameters, which is part of why a routing argument that avoids excavation hardware altogether is attractive [5].

The problem the design exists to solve is documented rather than asserted. In parabolic flight at one sixth g, JSC-1A lunar simulant in an hourglass hopper flowed very slowly and in some runs stopped altogether, restarting only when the assembly was tapped, shaken, bounced and hammered [4]. JSC-1A contains no agglutinates, and the authors note that real lunar soil in hard vacuum is more cohesive still, consistent with the cohesion values measured from Apollo-era soil mechanics experiments [7]. Sample delivery, not acquisition, is where several earlier missions lost science: the Venera landers on Venus returned imagery but did not complete their sample deliveries to the onboard analysis instruments [11], and cohesive icy soil defeated part of the Phoenix Mars lander’s robotic arm delivery to its thermal and evolved-gas instruments, sticking to the scoop and to the instrument doors rather than falling free [10], [2]. Rotary drilling into planetary regolith carries a related problem that pneumatic transfer avoids entirely, premature bit wear or fracture from running outside the correct speed for the formation, since a gas jet has no cutting element to wear or break [6].

Mobility

None. The sampling head is carried to the surface by its host’s deployment mechanism and does not move afterwards.

Power and energy

No power figure is published. The system’s own argument is that total draw is negligible because a single valve opens for a few seconds [4].

Thermal

The sampling head and transfer tubing sit outside on the arm; the collection system is on the lander’s temperature-controlled mid-deck [12]. No operating or survival temperature limits are published.

Compute and avionics

None of its own beyond the pneumatics manifold and the break-beam sensors. The commanding and the imaging both belong to the host lander [12].

Autonomy

None. One command opens the manifold, and the sequence that follows is a physical process rather than a controlled one.

Communications

Through the lander.

Payload and instruments

A pair of infrared break-beam sensors at the collection container inlet captures high-speed data on material passing into the chamber, giving a non-image verification that sample arrived [13]. A camera images the collected sample, both to correlate against the break-beam data and to image dust coupons mounted on the back wall of the chamber.

Modes of operation

Collection, then verification, then sorting. On command the manifold releases gas down the arm to the head; the gas exits the nozzles and lofts material up the tube, arriving in under a second; the break-beam pair records the passage; the sample is caught on etched screens along a labyrinth path [13]. After imaging, a second burst enters by a secondary path, vents the finest fraction with the exhaust gas and separates the remainder about 1 mm either side of a dividing wall, coarse on one side and fine on the other, because some instruments want a size-selected sample.

Ground operations

The lander team deployed the Sample Access Arm; the sampling itself was one commanded event [12]. No planning cycle or tooling is published.

Technologies developed

The flight demonstrated that a pneumatic sampler works on the Moon, and it demonstrated something less expected about its robustness. The sampling hood was not flush with the surface, appearing from the video to be aggressively angled or standing off it, and it still collected several times the required volume [13]. The team’s comparison is that the result resembles off-nominal laboratory tests with the head several inches clear of the sampled surface, which is a qualitative match rather than a measured one, and no corresponding ground case is cited by name.

Two capabilities beyond acquisition were shown in the same event: verification of a delivered sample without an image, from break-beam timing alone, and particle size selection in the capture chamber using nothing but a second gas burst [13].

The unplanned result was lens cleaning. Pneumatic operation cleared accumulated dust and propellant residue from camera lenses, visible both in the capture chamber after sieving and in the collection video [13], [12].

What the flight does not establish is nominal performance. There is one acquisition, from a position that was not commanded and cannot be reconstructed, so the collected quantity cannot be attributed to a known standoff, and the builders’ assignment of technology readiness level 9 rests on that single off-nominal event [13].

The same team has proposed the pneumatic approach beyond the Moon: a 2021 New Frontiers-class concept study for a Ceres sample return builds its acquisition step around PlanetVac heads in the lander’s footpads, sampling through the same gas-driven principle Blue Ghost flew before returning to a fixed-price capsule design [3]. That concept depends on transferring a pneumatically captured sample into a sealed return canister, a step the study itself names as an emerging design still under development elsewhere, which is the same open problem the Blue Ghost flight left for a repeat attempt: what a pneumatic sampler does after capture, at scale, has not yet been shown.

References