VIPER

NASA/Helen Arase Vargas. Public domain (NASA / US government work).

Overview

VIPER is a four-wheeled, solar-powered rover built by NASA ARC with NASA JSC, intended to prospect for water ice at the lunar south pole [9]. Its two mission objectives are to characterize the distribution and physical state of polar water and other volatiles in cold traps and regolith, and to supply the data needed to evaluate the return from in-situ resource utilization in the polar regions. The measurement it exists to make is not available from orbit: orbital remote sensing constrains the presence of polar hydrogen but not its form, character or distribution in the subsurface at the resolution and viewpoint a rover provides [8].

The flight has changed twice. NASA ended the project in 2024 after removing VIPER from Astrobotic’s Griffin Mission One, then in September 2025 awarded a CLPS task order to Blue Origin for delivery on the second Blue Moon MK1 lander in late 2027 [3], [11]. The mission profile below is that assignment.

Specifications

Parameter	Value	Source
Roving mass	450 kg	[12]
Wheels	4, independently driven and steered, actuated suspension	[8], [12]
Steering actuator range	+/-50 deg	[12]
Suspension linkage range	+/-40 deg
Mobility actuator operating range	-45 to +110 C
Unpowered shadow survival, mobility hardware	at least 80 h
Drill depth	1 m, excavated in 10 cm increments	[8], [10]
Cameras	8 visible cameras	[8]
Prospecting instruments	NSS, NIRVSS and MSolo, operating continuously while driving

Mission profile

Parameter	Value	Source
Delivery	Blue Origin Blue Moon MK1, CLPS task order CS-7	[11]
Task order value	190 million dollars
Delivery date	late 2027
Landing site	Mons Mouton, lunar south pole	[10]
Mission duration	about 100 days	[8], [9], [10]
Primary mission area	about 5 by 4 km on Mons Mouton	[7]

The rover build was completed on 4 June 2024 after 192 integration steps and 464 components and subassemblies [3], [4]. The primary mission area, chosen to hold a landing site and a rover traverse, lies on the southern edge of Mons Mouton, a flat-topped mountain northwest of Nobile crater [7], [10].

Mobility

Four wheels are independently driven and independently steerable. All-wheel steering allows the vehicle to point arbitrarily while roving, which is used to hold the solar array on a near-horizon Sun while translating [9]. Active suspension changes vehicle ride height, allows traverse of comparatively large obstacles, and controls the load carried by each wheel; offset steering combined with active suspension improves driving in soft soil. Wheel pose and load control is also used directly as a geotechnical instrument. A quadrature pattern machined into the inboard face of each wheel rim lets the hazard camera images be used to estimate wheel sinkage, and aft and navigation camera images of the wheel tracks left behind support secondary inferences about sinkage and gross wheel slip over the distance covered by an image [1].

Independent lifting and placement of a wheel supports gaited extraction from soft regolith, where continued rotation under high slip excavates rather than propels. Four gaits were characterized in sink tank testing, named slither, ground hog, butterfly and inchworm [6], [10]. Wheel endurance testing to failure covered 40 km against the 20 km traverse goal.

The prospecting instruments run continuously while the rover drives, so traverse rate is set by the dwell time those measurements need rather than by what the mobility system can do [8]. Traverse rate is a survival constraint as well as a productivity one, because the rover has to reach a sunlit or communicating position before the local shadow closes over it [7].

Power and energy

Three solar arrays of approximately 1 m2 each are mounted on the port, starboard and aft surfaces rather than horizontally, because polar solar incidence is grazing sunlight just above the horizon [5], [9]. Each array can power the rover alone, with optimum power when the Sun illuminates two panels at a rear corner simultaneously [4]. The arrays sit outside the multi-layer insulation and were qualified across the full external temperature range for that reason. Batteries are housed in the lower chassis; capacity in watt-hours is not stated in the retrieved documentation.

Permanently shadowed regions receive no direct sunlight during the year and minimal scattered light from surrounding terrain [5], so excursions into them run on battery. The mobility hardware is qualified to survive at least 80 hours of shadow conditions with no power [12]; no rover-level excursion duration inside a permanently shadowed region is published.

Thermal

Lunar surface temperature at 85 degrees latitude ranges from a mean 182 K at local noon to a mean 61 K shortly before sunrise, and the coldest permanently shadowed crater is approximately 30 K [5]. The rover is designed to function from about 30 K to 300 K [4].

Thermally constrained electronics sit inside a warm box whose four sides are heat spreaders holding a nominal electronics temperature range [4]. The rover core is wrapped in multi-layer insulation to block conductive and radiative transfer, and heat-rejecting radiators run along the top deck. During periods without sunlight the surface segment hibernates on survival heaters, so traverses are planned to place the rover where the shadow it must sit through is shorter than it can survive [7].

Thermal vacuum testing ran to 15 K, with the mobility system, suspension, pan and tilt and navigation lights actuated at that temperature [3]. TVAC was completed 16 September 2024.

Headlights are carried because the rover must image and navigate terrain that receives no sunlight [4], [9].

Compute and avionics

Avionics are consolidated in an Integrated Avionics Unit inside the warm box; its late delivery, more than 16 months, held up warm box assembly and hence overall integration [4]. Rover Flight Software runs on radiation-hardened RAD750 and radiation-tolerant Aitech SP0-S avionics [9]. Processor clocks and memory sizing are not published.

Software is split across the link rather than layered onboard [9]. Rover Flight Software is built on the NASA Core Flight System and provides low-level hardware interfaces, basic mobility control, waypoint driving, odometry, basic error checking and device and payload services. Rover Ground Software runs on commodity desktop computing at mission control as an ensemble of ROS 2 nodes and performs navigation, mapping and generation of driver decision support data [6].

Eight visible cameras are carried [8], [9]. A stereo pair of monochrome navigation cameras on a pan and tilt gimbal covers the mid and far field, and monochrome hazard cameras around the chassis cover the near field for hazard identification, hazard avoidance and fault management; the same imagery also returns data on rover and surface interactions such as wheel tracks. An inertial measurement unit, a star tracker and joint encoders supply localization, attitude and body rate estimation; the IMU is an LN-200S reporting chassis linear acceleration and rotational velocity, from which the lunar gravity vector, and so the local slope angle and vehicle attitude, are deduced for geotechnical use [1].

Autonomy

Driving is by teleoperation using individual position commands set several meters ahead of the vehicle [9], with a target of about 4 m per command [6]. Onboard functions are localization, hazard detection through the hazard cameras, and fault management; route selection stays on the ground. Assessments made by the drive team during driving and science activities happen in timeframes that vary from seconds to minutes, against an Earth to Moon round-trip communication time of about 6 to 10 seconds [2]. That figure is the end-to-end time from a transmission leaving Earth to the response arriving, not the one-way light time.

Software slip estimators predict soil slippage during a drive, and a star tracker attitude hold function was tested for maintaining Earth communication lock while traversing [10]. The operations model is described by the project as a hybrid of human spaceflight operations and Mars rover operations [6].

Communications

Communication is direct to Earth in X-band, worked through the Deep Space Network [9]. There is no relay orbiter.

Link geometry constrains the traverse. Whether a Deep Space Network station is visible from a given rover position is computed by the horizon method against a merged terrain model, a station counting as available only when it stands more than 2 degrees above the locally computed horizon; the resulting communication maps are generated on the same two hour cadence as the illumination maps and are used to plan the route [7].

Payload and instruments

NSS (Neutron Spectrometer System) detects hydrogen within roughly the upper meter by epithermal neutron suppression, operating while driving to give a continuous hydrogen indication along the traverse. Its sensor head is mounted on the front of the upper frame [4]. It operates while the rover drives, giving a continuous hydrogen indication along the traverse [8].

TRIDENT (The Regolith and Ice Drill for Exploring New Terrain) is a percussive rotary drill that excavates the subsurface to 1 m in 10 cm increments while the rover is stationary, so that NIRVSS and MSolo can examine the cuttings directly [8]. Percussion is required because in-situ lunar soil relative density is about 65 percent in the top 15 cm and exceeds 90 percent below 30 cm, against a 65 to 75 percent practical limit for terrestrial field compaction even with heavy construction equipment [5]. Cuttings are lifted into a surface pile whose stratigraphy preserves depth order for imaging and sampling.

The drill’s own telemetry has been used to estimate rock strength in the field, and the published results disagree with themselves. Two abstracts from the same September 2023 campaign near Bishop, California, using the same TRIDENT engineering model on the same densely welded Bishop Tuff by the same method, drilling specific energy converted to unconfined compressive strength through a relation calibrated on lunar simulant cements, give 40 to 80 MPa [15] and 10 to 20 MPa [16]. Neither abstract explains the discrepancy and neither acknowledges the other, and both are abstracts rather than full papers. The strength of that rock should be treated as unresolved within a factor of four rather than taken from whichever abstract a reader finds first. The two agree on the weaker unit at the same site, a pumice airfall deposit at 2 to 5 MPa, and on the failure mode that matters for a percussive drill: in welded tuff the auger stalled on a torque spike at 48 cm, exceeding the motor current limit several times and requiring operator intervention [15].

MSolo (Mass Spectrometer Observing Lunar Operations) measures the composition of volatiles evolved from drill cuttings, separating water from other species and identifying contaminants including lander propellant residue. It is mounted in the lower chassis [4].

NIRVSS (Near-InfraRed Volatiles Spectrometer System) measures surface and cuttings composition spectrally over 1300 to 4000 nm, discriminating water ice from hydroxyl-bearing minerals both along the traverse and in drill cuttings, and carries a thermal radiometer [8].

Modes of operation

On rails. Traverse segments between science stations, with the instruments powered and taking data along the way but not held to the measurement requirements that govern a prospecting run [8].

Prospecting. Mapping inside a science station, an area dominated by a single predicted Ice Stability Region. Ice stability regions are defined from ice stability depth maps computed across the mission area, which separate the subsurface thermal environments in which water ice is stable at the surface, in the shallow subsurface, or only at depth [7]. NSS returns bulk hydrogen while NIRVSS and MSolo measure surface hydration, volatile composition and mineralogy, all while the rover drives [8].

Subsurface assay. Stationary drilling with TRIDENT and analysis of the cuttings by NIRVSS and MSolo [8]. Drill sites are pre-planned, and confirming or selecting drill sites during surface operations is a standing task of the science team [2].

Hibernation. Minimum power mode, holding temperatures on survival heaters through a shadow or a loss of signal the rover has been positioned to survive [7].

Ground operations

The Mission Operations Center and Mission Science Center are in adjacent buildings within the Multi-Mission Operations Center at NASA Ames [2], [4], with backrooms at NASA Johnson and at vendor facilities [10]. Rover driving, mission monitoring and systems execution are at Ames; rover systems and thermal at Johnson; MSolo instrument operations at Kennedy [6]. The driving position is staffed as a three-person Drive Team of Driver, Co-Driver and Navigator working alongside real-time science. The Real-time Science console sits with the drive team and a second console, Science Lead, was added beside it after the science operations group assessed the workload on a single real-time scientist; twelve further console positions are assigned to scientists in the Mission Science Center.

Planning is two-tiered. Traverses, science stations and activities are scheduled before launch; in flight, the science team guides traverse planning and confirms or selects drill sites as surface operations proceed [2]. Change proposals originate with the Mission Systems Planners, for example after an update to achievable rover speed, or with the Mission Science Center, which works the change with the planners and proposes it to Mission Science for approval.

Traverse planning is map-based. Shape-from-shading digital elevation models at 1 m per pixel, merged with the polar LOLA terrain, are used to compute Sun visibility against time, direct-to-Earth communication availability against time, ice stability depth, surface temperature and slope across the mission area, on a two hour cadence through the mission [7]. Boulder and crater hazard layers and the combined constraint layers are overlaid to choose a route that keeps Sun, communication and acceptable slope available together.

The science operations system was not built as a separate component to be joined to the mission operations system late; the science operations group instead treated integration with the mission operations and ground data system groups as a daily and weekly work practice from the 2019 design stage onward, which is what made a science role inside the Mission Operations Center possible at all [2].

Ground software is a purpose-built Ground Data System whose Build-9 was declared feature complete in May 2024, covering lunar transit housekeeping and payload checkouts, lunar surface housekeeping, on-rails driving, science stations, PSR entry and exit, and Safe Haven hibernation [4]. Artificial intelligence planning tools developed at Ames support traverse planning. Training and verification used RSIM, a Gazebo-based lunar surface rover simulator built for the mission [6]. By September 2024 the operations team had completed 700 hours of test and training and multiple runs of six of the seven engineering simulations required for lunar surface operations, with all ground software features and procedures ready [3].

Program status

NASA Headquarters decided in January 2024 that VIPER would no longer fly on Astrobotic’s Griffin Mission One, and the resulting delivery uncertainty led the agency to conclude the project [3]. Environmental qualification was completed anyway: vibration testing and post-test checkout on 5 July 2024, acoustic testing on 23 July 2024, and thermal vacuum on 16 September 2024. A Request For Information for alternative lunar missions was released on 9 August 2024 with responses due 2 September 2024 [3]. The rover entered long-term storage in February 2025 awaiting mission arrangements [3].

Under CLPS task order CS-7, awarded 19 September 2025, Blue Origin takes responsibility for the landing mission architecture, end-to-end payload integration and post-landing deployment [11]. The 100 day science mission is retained, and its science window is what sets the late 2027 landing date [10].

Technologies developed

Environmental qualification included sine and random vibration, acoustic testing at 138 dB, and thermal vacuum to 15 K with mechanisms actuated at temperature [3]. Wheel endurance testing to failure covered 40 km [10]. Sink tank testing characterized four alternative locomotion gaits for the low-cohesion soils expected inside permanently shadowed regions [6].

Two dedicated geotechnical experiments are defined for the mission: an assessment of vehicle performance on sloped highlands terrain during rails driving, and a measurement of the effect of volatiles on geotechnical properties, both built on wheel slip and sinkage estimated from mobility telemetry at 10 Hz and from camera imagery collected at roughly 5 m waypoint spacing [1].