Scarab

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

Overview

Scarab is a lunar prospecting rover built at Carnegie Mellon University in 2007 to carry a coring drill into permanently shadowed polar craters [1], [4]. The reference mission lands the rover on the floor of a permanently shadowed crater, where it traverses kilometers in darkness on radioisotope power, stopping to drill into the near subsurface. Its chassis reconfigures to stabilize the drill against the ground and to change ride height for driving on steep unconsolidated slopes.

Scarab was the flight-like carrier for the RESOLVE payload, the Regolith and Environment Science and Oxygen and Lunar Volatile Extraction instrument suite, and drove with the RESOLVE drill and instruments at the PISCES lunar analog site on Mauna Kea, Hawaii [4]. The vehicle line that followed it is NASA’s Resource Prospector, whose rover was designed and built at NASA JSC with Ames [12], and which was canceled before flight [5].

Specifications

Vehicle characteristics as built:

Parameter	Value
Mass	280 kg
Weight, Earth surface	2740 N
Weight, lunar surface	450 N
Footprint	roughly 1.5 by 1.5 m
Drill tower height, upright	2.2 m high stance, 1.6 m low stance
Average power available	100 W
Nominal power	175 to 200 W
Idle power	24 W
Locomotion speed	3 to 6 cm/s
Wheel diameter	65 cm
Track width	1.4 m
Wheelbase	0.8 to 1.4 m, 1.2 m nominal
Track to wheelbase aspect ratio	1:1.0 low, 1:1.2 nominal, 1:1.7 high stance
Center of gravity, planar	on geometric center
Center of gravity height	0.48 m low, 0.64 m nominal, 0.74 m high stance
Static pitch-over angle	56 degrees low, 43 degrees nominal, 30 degrees high stance
Static roll-over angle	61 degrees low, 53 degrees nominal, 49 degrees high stance
Maximum straddle	55 cm
Minimum straddle	belly contact
Approach and departure angle	105 degrees nominal stance
Breakover angle	115 degrees nominal stance
Belly clearance	30 cm, maintained while the body pitches

Center of gravity height and the static tip-over angles were measured empirically on a tilt table rather than computed [1]. Pitch-over refers to tipping forward or backward and roll-over to tipping sideways; on Scarab these are symmetrical.

Mission profile

Scarab did not fly. The reference mission it was sized for is a crater-floor survey on the order of 25 drill sites, with kilometers of traverse between them and 25 cores taken [1], [4], run as a serialized cycle of drive, charge batteries, drill, charge, repeat. Power was assumed to come from a radioisotope thermoelectric generator of the Advanced Stirling Radioisotope Generator class.

In field testing the vehicle completed a continuous traverse of 1 km, roughly the distance it would cover between drill sites on such a mission, and drove terrain including graded slopes and craters about 5 m across formed in loose mixed-grain slag [1]. It climbed rock obstacles of 40 to 50 cm.

Mobility

Four wheels, chosen to reduce degrees of freedom while giving a stable base; more than four wheels over-defines a surface. Steering is skid steering rather than explicit steering, selected for system simplicity and on the precedent of the Lunokhod rovers. The suspension is passive and kinematic rather than sprung, because at 3 to 6 cm/s with lunar gravity acting on the vehicle mass there are no dynamic loads worth damping [1].

Skid steering trades away the actuators of an explicit steering mechanism at the cost of high wheel slip during turns; wheel design research for another lunar rover found that diagonal grousers reduce point-turn sinkage without adding mass or volume [3]. Widening or enlarging the wheel also reduced sinkage, but not without a mass and volume cost [3].

Left and right rockers are connected by a differencing linkage that averages the two sides: as one side pitches up the other pitches down by the same amount relative to the body, and the body pitch angle is the mean of the two, so the body pitches considerably less than it otherwise would. A geared differential was rejected because the center of the vehicle is occupied by the drill workspace; a linkage running around that volume was used instead. To keep 30 cm of belly clearance as the body pitches, the underbody is scalloped, with the reference case one wheel on top of a 30 cm rock [1].

Active adjustability is layered on top of the passive differencing and is independent of it, so wheelbase and body height can be set separately on each side and the body can be leaned. The adjustment mechanism is a five-bar doubler linkage driven by an electric linear actuator, similar to the linkages used to pitch backhoe buckets through large angles; a flight implementation would likely use a rotary actuator for sealing and stroke. Point turning is possible across a large range of poses with considerable wheelbase variation.

Each of the four wheels is directly and independently driven, with a brushless motor, planetary gearhead and harmonic drive embedded at the hub for a total reduction of 400:1 [1]. Rim pull per wheel was designed to approach the full weight of the vehicle.

The first-iteration wheel is a tubeless rubber skid-loader tire on a machined hub, which the design team identified as a terrestrial stand-in for a metal flight wheel [1]. The majority of suspension components are machined 7075 aluminum, and the differencing linkage is carbon fiber composite.

Power and energy

The rover is radioisotope powered [4]. Power on the field vehicle comes from an electrical simulator of an Advanced Stirling Radioisotope Generator supplied by NASA GRC, which supports untethered operation for periods of time [1]. The vehicle has 24 V and 48 V buses with distributed motor controllers. The power system was designed around a persistent, low-capacity source rather than a peaky one, which is why the concept of operations serializes driving, drilling and battery charging rather than overlapping them.

Thermal

Lunar surface rover concepts generally confront extreme temperature swings, dust and radiation alongside their terrain and delivery constraints [11]. A 2005 survey of exploration rover concepts concluded that modularity was the single most important factor for developing a viable surface mobility vehicle across missions [11]. The design environment is hard vacuum, 40 K ground temperature and a constant 3 K sky, with waste heat available from the radioisotope supply [1]. First-order thermal balance analysis indicated that the waste heat can hold electronics and other components in range but that excess heat must be shunted to a radiator surface, with a thermal management system to distribute heat and control temperatures and probably electrical heating of outboard components. Thermal paths to the drill and to any part of the vehicle holding icy soil have to be minimized so that volatile samples are preserved. The vehicle body is deliberately compact for thermal regulation [4].

Compute and avionics

A PC-104 computer with distributed motor controllers, wireless communication and wireless remote control. Much of the electronics is integrated to the chassis rather than to the body, which decouples it from the body structure and lets it dump heat into the central void of the rover.

Autonomy

Localization and navigation use wheel odometry, an inertial measurement unit, and terrain sensing by active devices, principally laser scanners, because there is no ambient light in a permanently shadowed crater [1], [4]. Wheel odometry and the inertial measurement unit estimate position and velocity in the dark. Flash and laser scanning systems were identified as the solutions that let the vehicle build terrain maps in permanent darkness. GPS was used for vehicle pose during early testing and afterwards as ground truth against which to evaluate the pose estimator.

Communications

Wireless communication and wireless remote control on the field vehicle [1]. No flight communications architecture is published.

Payload and instruments

Scarab is a drill carrier. The vehicle was sized around a drill of about 1 m length and about 3 cm diameter [4], and the mission requirement is 1 m geologic cores [1]. Drill thrust was expected to reach 100 to 200 N or more, and the vehicle has to supply enough weight on the lunar surface to react that thrust and the torque about the bit. That requirement, not the mobility requirement, set the vehicle mass.

Drilling for buried volatiles carries a loss risk of its own: thermal vacuum tests of the RESOLVE-lineage Honeybee Robotics Lunar Prospecting Drill found that roughly 30 percent of a sample’s water content was lost to sublimation during drilling and sample transfer, with loss decreasing at lower ambient temperature [7]. Moisture retention also correlated with sample size, with a larger sample losing proportionally less water [7].

The drill is mounted in line with the vehicle center of gravity, to maximize the weight available to react thrust, and bolted directly to a high-stiffness, high-strength chassis to maximize platform stability against thrust loads, torques and vibration [1]. The chassis is an aluminum weldment spanning between the suspension pairs, hollow at its center so that it wraps around the drill workspace. The structure supporting the drill continues upwards and doubles as the navigation sensor mast [4], a dual use that follows from the drill needing to be upright at a sampling site and the navigation mast needing to be upright while driving. A fixed drill structure was chosen over a deployable one for stiffness and to remove an actuation; a flight vehicle might need a single-use deployment depending on stowage.

The drill and the payload were external contributions: a breadboard coring drill and payload guidance from the RESOLVE team at NASA JSC and the Northern Centre for Advanced Technology, and rover sensing and navigation techniques co-developed with NASA Ames [1]. When the drill was not available Scarab carried a static stand-in replicating its mass properties.

Modes of operation

Three ride-height poses are defined: low, nominal and high [1]. Low stance lowers the drill to the surface and gives the best static stability, 56 degrees pitch-over and 61 degrees roll-over, and high stance gives the most clearance at the cost of stability, 30 degrees pitch-over and 49 degrees roll-over. The mission cycle is drive, charge, drill, charge, repeat [4].

Technologies developed

The design result that outlived the vehicle is the drill-driven chassis: a passive differenced rocker suspension for terrain conformance, with an independent active adjustment that changes ride height and wheelbase so the same structure that drives can be reconfigured into a stiff, low, heavily loaded drilling platform. The same combination of passive suspension with active adjustability had been used on NOMAD [9], and the resulting kinematics turned out very similar to NASA JPL Sample Return Rover, at a different scale [1]. NOMAD’s own most visible result was the first meteorite discoveries made by a robot, during a 2000 expedition to Elephant Moraine, Antarctica [9].

Resource Prospector

NASA’s Resource Prospector carried the same payload lineage forward. Its science premise was that LCROSS had confirmed water ice in a permanently shadowed region and indicated enough of it to be economically viable, and that RP would map the distribution at human scales of meters to kilometers and test whether areas of temporary sunlight also retain ice [6]. RP’s stated purpose was to ground-truth remotely sensed lunar water detections with a rover, drill and the RESOLVE science package, since only remote detection data existed before it [7]. RP was managed at Ames under the Advanced Exploration Systems division of the Human Exploration and Operations Mission Directorate, as a Class D, Category 3 mission [8]. The mission concept had been publicly introduced since at least 2014 [5]. RESOLVE became the RP payload, and the work was split across three centers: the rover at NASA JSC, the project and mission operations lead at Ames, and payload integration and functional testing at NASA KSC, where most of the instrument suite was first mounted to a ground interface structure that replicated the rover payload bay before shipment to Johnson [12].

Published RP rover parameters [8]:

Parameter	Value
Mass, including payload	300 kg
Dimensions	1.4 x 1.4 x 2 m, mast included
Speed made good	0.5 cm/s
Power	300 W, solar
Communications	direct to Earth
Surface duration	6 to 14 Earth days
Launch vehicle and date, as planned	Falcon 9 v1.1, 2020

The prototype built in 2014 and 2015 is a solar-powered four-wheeled vehicle with hub motor drive, offset four-wheel steering and active suspension [8]. Active suspension changes ride height, lets the vehicle cross comparatively large obstacles and controls wheel loading; all-wheel steering lets the vehicle point arbitrarily while roving, for instance to hold the solar array on the sun while moving; and offset steering combined with active suspension improves driving in soft soil.

The mobility requirement came from the volatile distribution. The top 0.5 m is hypothesized to be patchy at 100 m scales because of impact gardening, and the mixing timescale increases with depth, so more mobility reduces the required sampling depth [8]. Minimum success was measurements from two places at least 100 m apart, which may require a 300 to 400 m path length; full success was two places at least 1000 m apart, needing a 3000 to 4000 m path, including measurement and sampling in a permanently shadowed area and an ISRU demonstration. About 60 to 70 craters of 10 m diameter form per square kilometer per billion years, and sampling at least ten of them was expected to need 2 to 3 km of traverse [8].

The rover payload was the Neutron Spectrometer System for resource localization, the Near Infrared Volatiles Spectrometer System for sample evaluation, the Lunar Advanced Volatile Analysis instrument for volatile content, the Oxygen and Volatile Extraction Node for extraction, and a drill and auger for subsurface sample collection, with a mast carrying communications antennae and navigation cameras and radiators for heat rejection [8].

RP flight software was built on NASA’s Core Flight Software: the core Flight Executive, an Operating System Abstraction Layer and cFE-compliant applications, chosen so that existing spacecraft functions could be reused while rover-specific applications were written against them [12]. The 2015 build 1 provided hardware interfaces, basic mobility, waypoint driving, odometry and inertial measurement unit localization, basic error checking and camera services [8]. The RP15 testbed was taken from concept to a rover driving in the lunar analog rock yard at Johnson within a single year. Operations were deliberately distributed rather than co-located, with each team working from its home center: driver and co-driver and the mission command position at Ames, rover systems at Johnson, and the payload at Kennedy, linked by a three-center operations platform with shared displays, CCSDS gateways and voice over IP commanding.

NASA KSC designed and fabricated the launch-lock hold-down mechanism that would have attached the rover to the lander through launch and transit and released it after touchdown, with an engineering development unit built and tested at KSC; deployment was to be down a ramp from the lander [10].

Prototype development matured RP subsystems to TRL 5, with radiation testing of avionics, thermal and thermal-vacuum testing of mechanisms and gravity-offload testing of mobility planned to follow [8]. The mission was later canceled. NASA’s next lunar volatile-prospecting rover mission, VIPER, carries the same goal forward with a rover built by a different team, to map ice stability regions from meters to kilometers across the lunar south pole [2].