TENACIOUS

ispace, inc.

Overview

A rover of about 5 kg [1] has to do three things on the Moon: survive the lander’s descent and deployment, drive on regolith it can barely push through, and get through a lunar day on a body-mounted panel and a small battery. At that mass the mobility, thermal and communications budgets are set mostly by what the host lander provides, so the rover’s fate follows the lander’s.

TENACIOUS is a four-wheeled lunar micro rover designed, manufactured and assembled by ispace-EUROPE in Luxembourg, carried on the ispace RESILIENCE lander as the surface element of HAKUTO-R Mission 2 [1], [2]. Development was co-funded by the Luxembourg Space Agency through a European Space Agency contract under the national LuxIMPULSE program [2]. Its task was to drive around the landing site, image it, and scoop lunar regolith whose ownership would then transfer to NASA [3].

RESILIENCE launched 15 January 2025 and made a hard landing at Mare Frigoris on 5 June 2025 [1], [4]. TENACIOUS was never deployed.

No peer-reviewed or conference publication on this vehicle was located. Every vehicle number below comes from ispace material [1], [2]. Mission figures come from ispace and from reporting of the ispace failure investigation [4], [5].

Specifications

Parameter	Value	Source
Mass	approximately 5 kg	manufacturer [1], [2]
Length	54 cm	manufacturer [2]
Width	31.5 cm	manufacturer [1], [2]
Height	26 cm
Payload capacity	maximum 1 kg	manufacturer [1]
Structure	carbon fiber reinforced plastic frame	manufacturer [2]
Wheels	four, profiled for stable travel on regolith
Camera	forward-mounted HD camera
Sampling	shovel for surface regolith
Command and data path	via the lander to mission control

The CFRP frame is specified against launch and transit vibration rather than against surface loads [1]. Drive speed, slope limit, obstacle height, power, battery capacity, thermal limits, processor and radio parameters are not published.

Where the design sits among lunar micro rovers

Small lunar surface vehicles have converged on a few families, and TENACIOUS belongs to the conventional one. What has reached the Moon is narrow.

• Hopping and single-wheel rovers. JAXA’s LEV-1 separated from the SLIM lander during its final free fall on 20 January 2024 and performed the first hopping locomotion by a robot on the lunar surface, six hops with the direction chosen onboard from camera imagery [10]. A companion article by the LEV system lead records seven attitude-heading-hop cycles and a LEV-2 that located and photographed SLIM [11]. Both rovers were released by the lander rather than lowered from it [12]. This is flown evidence for ultra-small rovers, and its mobility is the opposite of TENACIOUS’s wheeled drive.
• Wheeled micro rovers on landers that failed. Rashid-1 flew on HAKUTO-R Mission 1 and never operated [6]. Its landing-site study used orbital slope and rock statistics only, with no ground truth [22]. Its two CASPEX cameras were calibrated before launch and returned no lunar image [21]. Of the wheeled micro rovers in this family, none has driven on the Moon, TENACIOUS included.
• Impact-tolerant micro rovers. The Dymon YAOKI is 15 by 15 by 10 cm and weighs 498 g, and the supplier states it withstands 100 G so that it can be dropped or tumbled into lunar caves [14]. The rating is a supplier claim with no stated test standard.
• Planned and studied wheeled rovers. The Canadian Lunar Rover Mission was first described as a 30 kg micro rover with an instrument set later replaced [24], and its status after preliminary design review gives a mass and six instruments but no mobility figure [23]. The European Moon Rover System Pre-Phase A study sets out a modular rover for four candidate missions, with only a terrestrial prototype field-tested [35]. Endurance-A is a concept for a long-range South Pole-Aitken traverse and sample return rover, and its figures are study values [34].
• Coverage planning for micro rovers. A myopic three-dimensional coverage planner fusing orbital elevation data with onboard LiDAR was validated in simulation on an Apollo 14 elevation model and in outdoor tests on a terrestrial rover [25]. It is not flown.

Mission profile

These describe the lander mission rather than the rover.

Parameter	Value	Source
Lander	ispace RESILIENCE	[1]
Lander payload capacity	up to 30 kg	[1]
Launch	15 January 2025, Florida, Falcon 9	[1]
Transit	low-energy trajectory, about five months	[4]
Landing attempt	5 June 2025, Mare Frigoris	[1], [4]
Success milestones achieved	8 of 10	[1]
Impact velocity	42 m/s	[5]

Descent proceeded from about 100 km to 20 km altitude with the main engine firing as planned, and telemetry was lost less than two minutes before the scheduled touchdown [4]. The failure investigation attributed the hard landing to the laser range finder, which was slow to return valid distance measurements, so the lander did not decelerate to the required landing speed and struck the surface at 42 m/s [5]. Installation error was considered and judged unlikely, since no anomaly appeared in testing or in the earlier descent phases, which left in-flight performance deterioration of the range finder as the leading explanation [5]. The mission achieved 8 of its 10 declared success milestones, missing only the landing itself and post-landing system stability [1].

RESILIENCE was ispace’s second lunar landing attempt. Both ended in an altitude sensing failure, by different mechanisms. Mission 1 flew its deceleration profile correctly to under 1 m/s vertical speed while its filter rejected a correct altimeter reading as faulty after a crater rim crossing made the sensed altitude jump, so the lander coasted on a stale zero-altitude estimate until it ran out of propellant about 5 km up and fell, and its payloads, including the UAE’s Rashid rover, were never operated [6]. Mission 2’s failure was a laser range finder slow to return valid readings during descent, so the two failures share an outcome and a sensing modality, not a cause [5].

Other commercial landers show the same dependence of a small payload on touchdown attitude. Intuitive Machines’ IM-2 touched down on 6 March 2025 about 250 m from its intended site, on its side inside a crater near the south pole, and was declared over the next day [29]. The Blue Ghost Mission 1 lander at Mare Crisium, in contrast, ran the Lunar PlanetVac pneumatic sampler through a five-second acquisition sequence [28].

Landing precision and the SLIM comparison

SLIM is a separate Japanese lander with no ispace involvement. Its flight record gives fourteen vision-based navigation fixes from lunar orbit through the vertical descent and a 3.4 m position error at the first 50 m hover image, so precision landing was evaluated as 10 m or better against a 100 m requirement [8]. One of its two main engines lost most of its thrust during the descent, and the cause is not diagnosed in that record [8]. The vision-based navigation and 50 m obstacle detection were validated by Monte Carlo and hardware-in-the-loop work before the single flight [9]. The mass reduction side of the project used a propellant tank that doubles as primary structure, thin film solar cells bonded to it, and crushable porous metal legs for a deliberate two-step tip-over landing [7].

These are one flight each, so they do not establish a distribution. They do show the contrast with RESILIENCE: SLIM landed with a navigation chain that measured against terrain features, and ispace’s two losses were both traced to altitude measurement. TENACIOUS never reached a surface on which its own design choices could be tested.

Sampling and the regolith contract

The regolith scoop exists to execute a contract rather than a science investigation. NASA selected four companies to collect lunar material and transfer ownership to the agency: Lunar Outpost at $1, ispace Japan at $5,000, ispace Europe at $5,000, and Masten Space Systems at $15,000, totaling $25,001 [3]. Payment is 10 percent on award, 10 percent on launch, and 80 percent on successful completion. Ownership transfers in place on the surface after the company sends imagery of the collection and data identifying its location, with no return to Earth.

TENACIOUS carried a shovel to gather surface material and the forward HD camera to photograph it [2]. The rover never needed to return or store a sample: the contract is satisfied by collecting, imaging and locating it.

Other sampling approaches in the corpus work differently. PlanetVac acquires regolith pneumatically with no arm or scoop and has flown once, at one site, from a head position the team could not control [28]. Neither approach is demonstrated on a micro rover.

Mobility

Four wheels, rigid, with no suspension in the published configuration [1], [2]. The wheels are described only as shaped so that the rover traverses regolith stably, and the flight model photograph shows rigid spoked rims with slotted circumferential grousers. Speed, slope limit and obstacle capability are not published, and no mobility test results have been released for TENACIOUS itself.

Grouser wheels of this class have been studied on Earth by several groups, and the results bound what to expect. A single-wheel testbed and discrete element model of a Rashid-1 style 100 mm radius grousered wheel found that traction coefficient and tractive efficiency at the flight load in 1 g sand changed little when the validated model was extrapolated to lunar gravity, while tractive force and resistive torque both dropped by about a sixth [15]. Every lunar gravity number there is model output, not measurement. A bench comparison of a 200 mm Rashid flight-design wheel in standard sand and FJS-1 simulant covered slip ratios to 50 percent, but it reduced load rather than gravity [16]. At translational speeds up to 1.0 m/s, about a hundred times flown rover speeds, traction depended strongly on speed, in a way low speed testing had missed [17].

Turning is the other limit for small skid-steer vehicles. Single-wheel tests in GRC-1 simulant found that wider (60 mm) and larger diameter (200 mm) wheels turned best and that diagonal grousers reduced point-turn sinkage [18], and lug alignment at 45 degrees reduced slip in a single-wheel crawler rig [19]. A variable grouser-height wheel, 0 to 17.5 mm, was evaluated in 750 testbed trials across gravel and sand, with the limit that wheel geometry and load were fixed [20]. LEV-1’s octagonal grouser was chosen with a resistive force theory model and then confirmed by operation on the Moon [13]. A Chrono-based continuum representation model reproduced single-wheel force curves for a 137 kg Yutu-class model, as a simulation only [36]. None of these results were measured on TENACIOUS or its wheels.

Astrobotic’s CubeRover line is the nearest American product class. It ran over 150 mobility tests with eleven wheel sets in a 120 ton regolith simulant bin at NASA Kennedy in December 2020, with some sets climbing 30 degree slopes [30], [33]. The vendor’s sheet lists 4.6, 8 and 10.6 kg rover masses for the 2U, 4U and 6U sizes, an offered specification rather than a demonstrated result [31], [32]. No CubeRover has driven on the Moon.

Power and energy

A body-mounted solar panel closes the top of the chassis, visible in the photograph above. Generation capacity, battery capacity and night survival are not published. The mission was planned inside a single lunar day, which ended at the landing attempt [1], [4]. A battery developed for the CADRE lunar rovers, a Li-ion cell selected for wide-temperature operation and qualified by vibration and thermal vacuum, shows what a rover of this scale requires, though it is a different vehicle [27].

Thermal

The chassis is wrapped in multi-layer insulation over the CFRP frame [2]. No temperature limits, radiator area or heater strategy have been published. Lunar dust is the main uncertainty for a vehicle that must also keep a body-mounted panel clean: solar absorptivity can rise by as much as 50 percent at only 12 percent area coverage [26]. That analysis targets the Lunar Terrain Vehicle and is not a TENACIOUS result. The YAOKI supplier reports replacing an aluminum body with cyanate ester CFRP, as TENACIOUS’s frame is also CFRP, with a stated mass saving of about 30 percent [14].

Compute and avionics

No processor, memory or radiation tolerance information is published.

Autonomy

None is claimed. Commands and data pass between mission control and the rover through the lander [2], which makes the vehicle a teleoperated one working inside lander communication range. LEV-1, by comparison, chose hop direction onboard [10].

Communications

Through the lander only [2]. Band, data rate and range are not published. LEV-1 transmitted directly to Earth on S band [10].

Payload and instruments

The forward HD camera is the only instrument named [2], alongside the regolith shovel [3]. Rover payload capacity is a maximum of 1 kg [1] against a 5 kg vehicle, so a hosted payload can be a fifth of the rover mass.

Ground operations

Operated from ispace mission control through the lander [2]. No planning cycle, tooling or operations concept has been published.

Technologies developed

The vehicle is the first European designed, manufactured and assembled lunar rover to fly [1], [2]. Because it was never deployed, no operational result exists. What the program established outside the hardware is a contractual path: a NASA contract that transfers ownership of collected material in place on the lunar surface [3].

What is not established

No independent test report, flight anomaly log or peer-reviewed publication on TENACIOUS has been located; every published vehicle number comes from ispace and ispace-EUROPE’s own mission material [1], [2]. Mass budget, drive speed, slope and obstacle limits, power generation and storage capacity, thermal survival range, processor, radiation tolerance, and communications band and data rate are all unpublished. No mobility, thermal vacuum or vibration test result for the flight unit has been released, and because the rover never deployed, none can now be generated by the mission. The regolith transfer contract that motivated the shovel and camera has not been executed by any of the four awardees in the sources held [3], [6]. The Mission 2 failure account rests on press reporting of an ispace statement, and the underlying ispace technical release is not held [5].