TRIDENT

Source: [1], CC BY 4.0.

Overview

A lunar polar volatiles mission has to answer a question orbital and impact data cannot: whether there is water at a specific spot, at what depth, and in what form, without melting or losing it in the act of finding out. Orbital neutron and reflectance data bound the problem to within a few weight percent and a few tens of centimeters of depth, and set the detection thresholds and lateral resolution a landed instrument needs to close the gap [3]. Closing it means putting a bit into permanently shadowed, cryogenic regolith and bringing the released volatiles to an instrument fast enough that sublimation has not already scattered them, which is a harder problem than excavating dry equatorial soil: the material may be indurated by ice cement rather than loose, cold enough to embrittle seals and gum lubricants, and valuable enough that the sample cannot simply be thrown away to get through it.

TRIDENT is the 1 m class rotary-percussive drill built by Honeybee Robotics to do this, on two lunar south pole missions: the PRIME-1 payload on the Intuitive Machines IM-2 Athena lander, and the VIPER rover [1], [14]. Two identical flight units were delivered, differing only in a set of calibration targets for VIPER’s near-infrared spectrometer. The drill does not analyze anything itself; its product is a cuttings cone deposited on the surface for a mass spectrometer, and on VIPER a spectrometer, to read, plus the telemetry the act of drilling generates [2].

Design space

The design descends from two decades of Honeybee drills built to the same general problem on different bodies: the Construction and Resource Utilization eXplorer (CRUX) concept paired a borehole neutron probe and evolved-gas analyzer with a drill to map subsurface hydrogen directly rather than infer it from surface neutron counts alone [5]; Icebreaker and the Life in the Atacama drill worked the Mars case, where ice cemented ground had to be penetrated without melting the very ice being sought [4]; and Resource Prospector carried the lineage back to the Moon before its 2018 cancellation. Mass fell from over 50 kg at Icebreaker to 15 kg at the Life in the Atacama drill before rising again for the flight configuration, as the requirement moved from a rover-mounted demonstrator to a lander- and rover-qualified flight instrument [8].

Two choices distinguish TRIDENT from that lineage and from a conventional auger. Rotation and percussion are driven by separate actuators, so the drill can run in rotation alone, percussion alone, or both, and spend the percussor’s power only when the material needs it [1], [2]. And the auger removes material in bites rather than boring one continuous hole: after about 10 cm of penetration it is withdrawn, a passive brush scrapes cuttings off, and a chute deposits them on the surface as a cone, before the drill re-enters the hole for the next 10 cm [8]. Bite sampling was chosen specifically to reduce the risk of freezing in on a re-entry into an empty hole, at the cost of an operation that pauses every 10 cm rather than running continuously [2]. Field testing shows what that choice buys and what it does not: at Haughton Crater and on the Bishop Tuff, TRIDENT penetrated unconsolidated and cuttings-permeable ground easily but choked and bound in low-permeability microporous targets, exceeding safety torque limits and requiring a wrench to free the auger, a fault the bite-sampling logic does not by itself prevent [7]. The operational fix demonstrated in the field was to cut the bite size from the nominal 10 cm down to 1 to 2 cm so cuttings are bailed more often, which restored slow, unassisted progress [7].

TRIDENT flew on IM-2, launched 27 February 2025 and landed 6 March at Mons Mouton [14]. Athena came to rest on its side, so no hole was drilled; the drill was instead run through its full range of motion in a horizontal orientation while buried in regolith thrown up during landing. VIPER, delayed by cost and schedule growth documented in its own program reviews, had not flown as of this writing [6].

Specifications

Parameter	Value	Source
Total mass	26.5 kg	[1]
Drill hardware mass	19.5 kg	[8]
Avionics mass	7 kg	[8]
Height	approximately 1.75 m	[1]
Bit diameter	25.4 mm, percussive-grade carbide tip	[2]
Feed stroke	1 m	[1], [2]
Deploy stage stroke	0.35 m	[1]
Percussive blow energy	2 J at up to 972 blows per minute	[2]
Percussor rated power	300 W	[8]
Nominal rotary power	50 W, 87 W average auger power	[1], [2]
Auger speed	100 rpm nominal, 10 to 120 rpm commandable	[1]
Feed speed	1.25 mm/s drilling, 2 mm/s otherwise, 0 to 4 mm/s commandable	[1]
Weight on bit	100 N nominal, 0 to 500 N commandable	[1]
Linear stage rating	500 N bidirectional	[2], [8]
Percussion rate	729 blows per minute nominal, 0 to 1000 commandable	[1]
Bite length	10 cm nominal, 1 to 10 cm commandable	[1], [8]
Cuttings per bite	about 12 cc per 10 cm	[8]
Actuators	four brushless DC, with planetary, spur and bevel gearing	[8]

Mission profile

Parameter	Value	Source
Launch	27 February 2025, 00:16:30 UTC	[1], [14]
Landing	6 March 2025, about 17:30 UTC	[1], [14]
Landing site	Mons Mouton, 84.7906 S, 29.1957 E	[1], [14]
Lander attitude	on its side	[1]
Drill activation	about 20 minutes	[1]
Deploy stage extension achieved	35 cm	[1]
Feed stage extension achieved	100 cm	[1]
Regolith penetrated	none	[1]

With the lander horizontal there was still enough power to exercise the drill. Over about twenty minutes both rotation and percussion actuators were run, both launch locks released, both linear stages driven to their full extension, the auger heater cycled, and the two drill string temperature sensors read out, all while the mechanism and avionics box were under a thick layer of regolith deposited during landing [1], [14]. The builders’ assessment is that the system operated within flight parameters and behaved as it had in preflight test.

VIPER’s assignment is different in kind: about three drilling deployments at each science station across a 100 Earth-day nominal mission in a 5 by 5 km area on Mons Mouton, with drill sites chosen tactically by the science team from the neutron spectrometer’s readings [1]. The lateral traverse VIPER adds is itself a response to how coarse orbital detection is: a rover that can range across a permanently shadowed region and stop wherever the subsurface neutron signal warrants is a different, and more capable, instrument than a single fixed drill [3].

Approach

Two nested linear stages carry the drill. The deploy stage presses a footpad against the surface for stability, and the feed stage advances the string. Both use a steel wire rope capstan on a rolling carriage rather than a ball screw, chosen because lunar regolith is more abrasive than Martian dust and a ball screw fails if dust enters it [8]. The capstan also lets the drill head float during percussion, which is the mechanical analogue of not pushing down on a jackhammer.

The auger is built in two sections to serve the bite-sampling scheme. The lower sampling section has deep flutes so material jams in and stays there; the longer upper section has shallow flutes for conveyance [8]. The combination is deliberately efficient at sampling and inefficient at conveying, so drilling a full meter in one run raises torque and power and increases the chance of a cuttings jam. Three consequences follow from bite sampling itself [8]: the cuttings retain a 10 cm stratigraphy instead of being mixed over a meter; drilling torque stays low, because cuttings are never conveyed the full height of the hole, and on re-entry to an empty hole the initial torque can be attributed to material strength alone rather than to strength plus chip transport; and the bit has time to cool between bites, which matters for retaining volatiles in the sample.

Mobility

None. TRIDENT is fixed to its host and reaches only along its own feed axis.

Power and energy

Nominal rotary drilling power is 50 W to 87 W average; the percussor is rated at 300 W [1], [2], [8]. No battery is carried and no energy budget for a complete hole is published. On PRIME-1 the entire activation ran on power the toppled lander could still supply.

Thermal

Kapton film heaters warm the avionics and actuators to operating temperature, and a heater inside the auger with a colocated temperature sensor 35 cm above the bit serves three purposes: thermal conductivity measurement, freeing the string if ice adheres to it, and driving sublimation from the cuttings pile [9]. Flight units passed thermal vacuum, vibration, shock and electromagnetic compatibility testing. Bit temperature during drilling is itself a measurement rather than a constraint: in engineering unit tests the bit cooled at visibly different rates in dry and in icy simulant, because dry lunar soil is insulating and the stem retains more heat in it [1].

Compute and avionics

The avionics box holds four motor drivers, a thermal card, a power distribution unit and a command and data handling board [9]. Onboard software commutates the brushless motors and strings low-level actuator commands into named operations: release launch locks, calibrate actuators, deploy the footpad, drill, retract to deposit cuttings, stow. No processor part or radiation tolerance approach is published. A rotary slip ring carries power and data to the auger heater and its two platinum resistance thermometers, and flex harnesses carry power and data across the full travel of both stages.

Autonomy

The drill is operated in near real time rather than autonomously, a decision the builders justify as simplifying avionics development while letting operators catch slippage, clogging and unanticipated behavior as it happens [2]. One behavior is onboard: percussion is normally triggered automatically when feedback shows the drill is making insufficient progress. Everything else is commanded.

Communications

Through the host vehicle. No band, rate or latency figure is published for the drill itself.

Payload and instruments

TRIDENT carries two platinum resistance temperature detectors, one integrated in the drill bit and one in the auger beside the heater, and no other instrument [8]. Its scientific output is otherwise indirect: the cuttings cone it builds is what MSolo and, on VIPER, NIRVSS measure, and its own telemetry is a geotechnical instrument. MSolo is itself a modified commercial residual gas analyzer, ruggedized for flight and qualified through vibration, thermal vacuum and EMI/EMC testing before delivery, with its ground-measured limit of detection and isotope-ratio reproducibility deferred to lunar performance that has not yet been measured [10].

Three derived measurements come out of drill telemetry. Specific drilling energy relates to unconfined compressive strength through an empirical baseline fitted across rocks from 4 to 120 MPa, and rate of penetration varies inversely with the same quantity [1]. Cone index follows from weight on bit and penetration rate before drilling starts. A modified footpad deployment, driven under current limiting with position telemetry read back, yields a force-penetration curve and from it the modulus of subgrade reaction and ultimate bearing capacity, quantities used in the Apollo era to predict how far surface packages would sink [11]. Those Apollo-era relations were built from six equatorial sites and carry their own stated caveat that a footing loaded outside the stress range they were fitted in is being extrapolated [11], a caveat that applies directly to a polar footpad reading taken at loads and temperatures Apollo never measured. A more recent synthesis of the same Apollo and Soviet dataset gives closed-form density and shear-strength-with-depth curves an excavator or driller can size against, while noting that almost none of the underlying data is from polar terrain [12].

A thermal vacuum campaign with the engineering unit and a flight MSolo unit drilled four 1 m holes into a 100 kg bin of lunar highlands simulant doped to 0, 2.5 and 5 weight percent water, at 1.6e-6 Torr and about -178 C, and showed that a mass spectrometer watching the cuttings pile can distinguish the doping levels by the water signal, with peak intensity correlating to a water fluence computed from the pile’s own geometry [9]. Pile dynamics, cuttings falling back into the hole or spilling past the bin edge, produced large spurious spikes and complicated that reading, and only one of the four holes had its pile geometry measured directly, so the fluence correlation is fitted to a single hole [9].

A separate Antarctic field campaign with the engineering unit showed that drill telemetry alone identifies subsurface horizons: an ice-rock transition at about 62 cm, a frozen boundary at 20 cm, and a bubble-rich ice layer at about 50 cm that matched published bubble content peaks [8]. Ice, soft soil and hard rock each produce a distinct signature in torque, weight on bit and penetration rate. Those results come from Antarctic ice and till at Earth gravity, not lunar regolith, and the fault detector built on them was tuned on the same data it was later scored against rather than validated blind [8].

The percussor has also been examined as a seismic source. A single 2 J blow produces an impulse above 50 Hz detectable 62 to 76 dB above the noise floor on accelerometers mounted on the drill chassis [9]. The flight inertial measurement units on both missions sample at 100 Hz against the 200 Hz used in the laboratory, so content above 50 Hz may not survive.

Ground testing

Source: [1], CC BY 4.0.

Flight units were environmentally tested but never used to drill simulant, and their drill strings were replaced after testing, so no flight-unit drilling data exists [1], [7]. All vacuum drilling performance comes from a dedicated engineering unit, which cut nearly 30 m in roughly 1 m intervals across 337 bites and more than 30 hours.

Test condition	Result	Source
NU-LHT-2M and LSP-2 simulants, below -100 C and below 1e-4 Torr	no measurable dust ingress effect on actuator no-load currents	[1]
LSP-2 doped to 8.5 wt% water, vacuum	bit cooling rate distinguishes icy from dry soil	[1]
NU-LHT-3M doped 0, 2.5, 5 wt% water, 1.6e-6 Torr, -178 C, with flight MSolo	mass spectrometer distinguishes doping levels; pile geometry confounds later holes	[9]
Rock coupons, 4 to 120 MPa unconfined compressive strength	specific drilling energy and penetration rate baselines	[1]
Percussive drilling to failure	cam-spring life 2 to 3 days of continual percussion	[8]
Haughton Crater and Bishop Tuff field campaigns	choked and bound in low-permeability targets; bite size reduction restored progress	[7]
Antarctic field campaign, 15 boreholes	percussive head broke after two days, repaired from field spares	[8]

The icy simulants are not lunar ice. Qualitative volatile measurement was not a baselined requirement on either mission, so no complex ice-formation method was used and the samples are water-doped simulant rather than an analogue of the polar deposits the drill was built to sample [9]. Reduced-gravity strength testing of terrestrial sand, run on parabolic flight with only two to three data points at lunar gravity, finds that residual friction angle and dilatancy shift with gravity in ways a 1 g or vacuum-chamber-at-1 g drilling test cannot reproduce, which bears on how far any of the ground campaigns above stand in for lunar surface conditions even where vacuum and temperature are matched [13].

Modes of operation

Rotation, percussion, and rotation with percussion, selected by the operator or triggered onboard when progress stalls [2]. Between these sit the named sequences the flight software carries: launch lock release, actuator calibration, footpad deployment, drilling, retraction with brushing, and stow.

Ground operations

Three console roles: a payload uplink lead who owns execution and sends every command, a payload downlink lead who watches telemetry and reports instrument health, and a TRIDENT Science role that represents the drill in the science center, recommends drill sites and advises when conditions warrant deviating from the procedure [1]. On VIPER the uplink and downlink leads sit in the payload operations control center and the science role in the mission science center; on PRIME-1, with fewer instruments and a slower pace, all three were colocated with MSolo. The ground software is a graphical interface named Poseidon that sets command parameters, sends commands, and displays telemetry as indicators, plots and event records; procedures are organized in a separate tool named PRIDE so that drilling can be synchronized with the instruments watching the cuttings pile. VIPER’s own ground test program before the 2022 program review exercised rover-side systems around the drill, not the drill itself: a 40 km wheel endurance run, slip characterization, and sink-tank tests in fine glass beads to develop escape gaits should the rover become entrapped near a science station [6].

Technologies developed

The bite-sampling architecture, and the pairing of a deep-fluted sampling auger with a shallow-fluted conveying auger that makes it work, is the transferable result, refined against the specific failure mode, ice-cemented and microporous ground choking a continuous auger, that stopped its Mars-drill predecessors [4], [7]. So is the capstan and rolling carriage feed, adopted specifically because a ball screw is vulnerable to abrasive dust, and the decoupling of rotation from percussion so that percussive power is spent only when the material demands it.

The empirical relations between drill telemetry and material properties are the second result, and they are what makes a drill a geotechnical instrument rather than only a sample acquisition device: specific drilling energy against compressive strength, rate of penetration against the same, bit cooling rate as a qualitative indicator of ice, and footpad force-penetration as a bearing capacity measurement [1]. The Antarctic campaign extended these into automated horizon detection from telemetry alone [8].

What is not established

Every penetration number attributed to TRIDENT is from a ground engineering unit in simulant or terrestrial rock, or from terrestrial field sites at Haughton Crater, Bishop Tuff and in Antarctica, not from lunar regolith [1], [7], [8]. PRIME-1’s flight activation exercised mechanisms while buried in regolith thrown up at landing, without drilling a hole, and VIPER has not flown. The icy simulants used in every vacuum test are water-doped analogs prepared without a complex ice-formation method, since qualitative volatile measurement was not a baselined requirement, and are not an analogue of the polar deposits the drill was built to sample [9]. The mass-spectrometer thermal vacuum campaign that comes closest to true polar conditions measured pile geometry, the quantity its water signal is calibrated against, on only one of its four holes [9]. The bearing capacity and shear strength relations used to interpret footpad telemetry, both the original Apollo-era values and the more recent synthesis of them, are built almost entirely from equatorial and mid-latitude sites and are explicitly not qualified for polar regolith [11], [12], and separate reduced-gravity testing shows soil strength parameters shift with gravity level in ways no drilling test at Earth gravity captures [13]. The Antarctic telemetry-based fault detector was tuned on the same campaign it was scored against rather than validated on independent data [8]. No flight drilling result exists for TRIDENT on either mission as of this writing.