Commercial Lunar Propellant Architecture - A Collaborative Study of Lunar Robotic Services - o5 Introduction

INTRODUCTION o5 - Robotic Services

For decades, humans have fantasized about living on the moon, but we haven't sent anyone there since 1972. So, how close are we to a moon base?

To the Memory of Dr. Spudis

Dr. Spudis earned his master’s degree from Brown University and his Ph.D. from Arizona State University in Geology with a focus on the Moon. His career included work at the US Geological Survey, NASA, John Hopkins University Applied Physics Laboratory, and the Lunar and Planetary Institute advocating for the exploration and the utilization of lunar resources. His work will continue to inspire and guide us all on our journey to the Moon.

“By going to the Moon we can learn how to extract what we need in space from what we find in space. Fundamentally that is a skill that any spacefaring civilization has to master. If you can learn to do that, you’ve got a skill that will allow you to go to Mars and beyond.”

Unified Geologic Map of the Moon

"People have always been fascinated by the moon and when we might return," said current USGS Director and former NASA astronaut Jim Reilly. “So, it’s wonderful to see USGS create a resource that can help NASA with their planning for future missions.”

The lunar map, called the “Unified Geologic Map of the Moon,” will serve as the definitive blueprint of the moon’s surface geology for future human missions and will be invaluable for the international scientific community, educators and the public-at-large. The digital map is available online now and shows the moon’s geology in incredible detail (1:5,000,000 scale).

Orthographic projections of the "Unified Geologic Map of the Moon" showing the geology of the Moon’s near side (left) and far side (right) with shaded topography from the Lunar Orbiter Laser Altimeter (LOLA). This geologic map is a synthesis of six Apollo-era regional geologic maps, updated based on data from recent satellite missions. It will serve as a reference for lunar science and future human missions to the Moon. Credit:

NASA/GSFC/USGS

Overview - Robotic Services

Robotic Services are crucial to any ISRU effort and need to be implemented correctly to ensure resilient, efficient operations. A fundamental ground rule of this study is the assumption of an uncrewed ISRU architecture to avoid the enormous burden of supporting crew and providing the associated crew infrastructure. Two traditional schools of thought bookend the spectrum of robotics in space; fully autonomous and completely teleoperated systems.

The optimal system for robotic services on the lunar surface is somewhere in the middle. For ISRU to be feasible in the short-term, robots will have to perform most tasks autonomously, communicate with each other, and work together toward a common objective, while being watched over by teleoperators in case intervention is needed. However, if these robotic systems are designed with requirements that exceed the current state of Artificial Intelligence (AI) technology here on Earth, then lunar ISRU operations will not get off the ground.

One of the most important robotic services that will be conducted on the Moon is prospecting, the identification and mapping of resources. This is further described in the Prospecting section.

The first robotic service is the delivery of the equipment that will be conducted with autonomous robotic landers. There is a multitude of companies working on this portion of the operations with large organizations and smaller startups making significant progress in the development of such technologies.

If the payloads from the lander are not able to dismount by themselves, the next robotic service is going to be an unloading service. To minimize additional equipment, the same robotic system used for unloading equipment from the lander should also be able to load equipment back onto the lander for ascent to orbit. Jet Propulsion Laboratory’s (JPL)’s All-Terrain Hex-Limbed Extra-Terrestrial Explorer (ATHLETE) 68 demonstrates a robotic concept that may be applicable to this task.

For the aforementioned techniques of extracting volatiles using surface and subsurface heating, mentioned in Mining Operations, robotic services to assemble/disassemble and move the tent structure would be crucial. Either the robots could be a part of the mining organization or a different robotic services company that partners with the mining organization. Due to the importance of these robotic services, backup solutions should be in place.

Robotic services for transportation of goods and materials are crucial to the ISRU effort as well and are described further in the Transportation section.

Due to the chemical and geometrical properties of lunar regolith, and the harsh environment of space, most systems will need some sort of maintenance to operate for extended periods. For this robotic service, the goal is to minimize mass while providing maintenance for a variety of different systems.

To address this challenge, some research and development may be required. Automated operation of the robots can be implemented for many tasks that are simple and well defined. Teleoperation should be limited to complex, infrequent tasks and the resolution of anomalies. Software updates can be introduced over time to increase efficiency of the operation while minimizing risk to other ISRU systems and operations. These types of robots could also have substantial uses here on Earth in the automotive, military, and servicing industries.

Universal Platform

Such robot platforms are currently in development for terrestrial applications and then for space mineral extraction, processing and construction. OffWorld 70 is a commercial company developing autonomous mining robots for Earth, Moon, Mars and asteroid applications with machine learning capability.

The common platform can be configured for multiple species as seen in Figure 28, though not all of these concepts are necessarily needed for every ISRU architecture. The assumption is that large numbers of smaller redundant robots protect against failures by offering many spares. The universal platform with modular attachments coupled with high production rates also drives costs down when compared to one-off highly custom robotic systems.

Figure 28: OffWorld Universal Robotic Platform

The architecture of the commercial lunar propellant system is being designed without a requirement for human presence. This, of course, means that all phases of the operation—construction, transport, maintenance, repair, extraction, refinement, propellant storage and transfer - must be executed by robotic means. Transport, extraction, refinement and propellant handling are the subjects of other sections. This section will discuss construction, maintenance and repair of the installation by robots.

Complex robotically assembled and maintained systems are a subject of much current progress and development. NASA, for example, is studying 72 how to assemble a very large telescope for exoplanet discovery and astrophysics missions. NASA is also supporting 73, 74 and 75 commercial development of robotic in-space assembly applications. Many of the considerations and capabilities for that in-space project have common elements with the use of robots on the lunar surface. Key questions that are being examined by NASA are:

• What is the right approach to modularizing the system, to achieve the highest performance and reliability at the lowest cost?
• What specific robotic tools and behaviors are required, and what is their technological maturity?
• How does the availability and cost of launch vehicles influence design choices?

Robotic construction is being developed by many research groups around the world. Entire houses are being built using concrete with 3D printing techniques. Lunar regolith could actually be used 76 in a similar way (Figure 29). Although many of the requirements for robotic construction, maintenance and repair will be for specialized components produced on Earth and launched to the lunar site, the use of sintered/printed regolith may prove very effective for structures like landing pads and ground transportation routes. However, in a permanently shadowed crater there is uncertainty in the effects that volatiles may have when mixed with the regolith being sintered.

Figure 29: Concept for Construction of Blast Shield and Roadway Using Regolith

Robotic assembly operations across the facility will include the following tasks:

• Landing site: minimal construction required, with blast shields being the primary requirement.
• Power plant: installation of solar PV farm, solar reflectors, and/or fission reactors; cable connections between plant components (collectors, control systems, power converters, radiators).
• Extraction facility: erection of the multi-component sublimation facility; integration of plumbing and/or transport for extractant; installation of control equipment.
• Processing and storage facility: emplacement of component equipment and tankage; electrical and fluid connections.

Facility components will be designed in such a way as to make robotic assembly convenient and reliable. Considerable laboratory work has shown the feasibility of robotic in-space assembly, such as for trusses (Figure 30). These concepts build on in-space construction experiments performed by astronauts (Figure 31). Industry has also made large investments in robotic assembly (such as Figure 32).

Figure 30: NINJAR 2.0, NASA Robotic Space Assembly Experiment 7

Figure 31: Assembly of the ACCESS Experiment during STS-61B 79

Figure 32: Robotic Construction of a Trestle Bridge

Some adjustments to robot and component designs will be required when leveraging ongoing work on in-orbit assembly. There are significant differences between robotic assembly in orbit and assembly on the lunar surface within polar craters. The key differences include:

Temperature. Temperatures within the polar craters are much lower than satellites on orbit experience. When satellites experience brief periods in eclipse, electric “survival heaters” are used to maintain components above minimum temperatures. Within the craters, however, shadow and very low temperatures are uninterrupted. Robotic mechanisms must be designed for these conditions. The use of liquid lubricants for joints and bearings, for example, will probably have to be changed to dry surfaces.

Abrasion. The Apollo astronauts brought back samples of lunar regolith - some of which got into their noses, sinuses and lungs. Regolith particles are sharp and abrasive. Moving parts of robots should be protected from dust intrusion.

Control. Robotic assembly schemes for on-orbit operations rely on teleoperation—Earth-based operators sending commands to the robots and monitoring progress with cameras. This requires nearly continuous communications, and proceeds slowly. The lunar facility may experience intermittent communications or limited bandwidth. Thus, a greater percentage of the operations should be automated in order that construction proceed efficiently.

Power. Satellites generally get their power from solar panels. This obviously will not work in the dark lunar craters. Power is discussed in detail in the Power section of this paper. Power can be provided to robotic systems there in four ways:

• Beaming power from the crater rim or other source
• Power provided in conductors contained in tracks or electric cable
• Battery storage in the robots
• Onboard RTG systems

Each of these has its advantages and disadvantages, but no major technical advances are required to power the robotic construction crew.

One approach to robotic operations is to install a track system. Elevated robotic tracks and rails would alleviate all of the risks described - temperature, abrasion, and control - and incorporate power delivery conductors. An elevated track allows for a more compact robot fuselage design and elimination of an articulated suspension, both of which would conserve heat inside the robot.

Moving above the regolith on a track, rather than directly on regolith, would lengthen the robotic lifetime considerably. Control of robot motion would be greatly simplified, as motion is one-dimensional and no obstacles need be avoided once the tracks are in place. Tracks could also have power conductors incorporated to power the robots directly, eliminating the need for power beaming or battery recharging.

Robotic maintenance and repair - both of the facilities and the robots themselves - will greatly benefit if a modular design approach is followed. Module replacement of robotic components should be very similar to facility construction, and hence not require additional techniques to be developed.

Highly dexterous operations and complex assembly should be avoided through design modularity. Adequate spare modules should be maintained on the site so that repairs can be made quickly, without serious interruptions to propellant production. Autonomous failure detection and response will be highly beneficial for efficiency and safety.

In Space

Grappling Arms

Robotic grappling arms would be a strong candidate for incorporation into a lunar orbital propellant depot or staging facility. Functions would include grappling and berthing of visiting vehicles, manipulation of on-orbit-replaceable modules, unanticipated repair needs, and external inspection.

NASA’s concept for the lunar orbiting Gateway currently includes robotics for such purposes. The Canadian Space Agency (CSA), for example, is sponsoring 82 work on the Next Generation Canadarm, or Canadarm 3 (Figure 33). Because the Gateway will only intermittently be occupied by humans, there will also be a need for internal robots, to perform maintenance and repair tasks when no humans are aboard.

Figure 33: Canadarm 3 on Lunar Orbiting Facility

The Near Rectilinear Halo Orbit (NRHO) node of the commercial propellant system may or may not be associated with the Gateway, but it will certainly have similar requirements for robotics. Using external robotic systems to capture and gently berth an arriving space object has significant benefits over hard docking: lighter interfaces, lower shock levels, and adaptability.

In some cases, the cargo might be transferred from the arriving vehicle to the NRHO facility without the need for berthing at all, but simply by using a robotic manipulator to reach out to the vehicle, grasp the cargo, and transfer it into the NRHO platform. A similar operation is performed by the Space Station Remote Manipulator System (SSRMS) on the ISS with vehicles that have external bays (Figure 34).

Figure 34: SSRMS on the ISS Reaching into the Dragon External Cargo Bay

A significant question to be considered in designing the space segment of the propellant architecture is whether it should include an escorting vehicle to assist in the capture of arriving objects. This is both a cost and a safety consideration. Today, there are a number of vehicles capable of rendezvous and docking such as the SpaceX 85 Dragon capsule (Figure 35).

The operations that are executed by these vehicles are complex, and many failure modes exist 86 . The vehicles are outfitted with a sophisticated suite of sensors that provide redundant information to the guidance system. Although the use of a reusable lander with rendezvous capability could amortize the cost of automated rendezvous and proximity operations over many ISRU deliveries, it would be very costly if every payload sent to the NRHO station had to carry these capabilities.

Figure 35: Dragon Capsule on Approach to the ISS - The SpaceX Dragon is illustrated on approach to the ISS, with parameters being continuously monitored

Alternatively, if there was a retrieval tug responsible for all sensing, thrusting, rendezvous and berthing operations, the complexity of equipment on cargo payloads could be greatly reduced, lowering the system- wide cost. A model for this “catcher” vehicle is Defense Advanced Research Projects Agency’s (DARPA)’s Robotic Servicing of Geosynchronous Satellites 88 (RSGS) vehicle (Figure 36).

A commercial variant of this model is also being developed by Altius Space Machines and described in the following Rendezvous and Capture section. Using these systems would allow resupply payloads to be “dumb,” as the “catcher” vehicle could remove orbit insertion errors, execute rendezvous and capture, and move the payload to the NRHO facility. In addition, robotic capture systems are inherently multi-purpose. Thus, the escort vehicle could potentially perform other functions such as external repair, assembly, module installation and removal, and close external inspection.

Figure 36: Illustration of the DARPA/SSL RSGS Vehicle - Shown are the two robotic manipulators, sensing suite, and other hardware

Rendezvous and Capture

In addition to the above-mentioned traditional in-space robotics approaches, several technologies are being developed for low-cost LEO satellite servicing that may also be directly relevant to rendezvous and capture operations at orbital transportation nodes in this propellant transfer architecture. ISRU only works if we have the capability to rendezvous so we can transfer propellant to customers.

In order to enable low-cost rendezvous and capture of LEO constellation spacecraft, Altius Space Machines 91 has been pioneering the use of boom-assisted magnetic capture technologies that allow for simplified capture of cooperative targets that are equipped with ferrous grappling targets. The use of a 6 Degrees of Freedom (DOF) robotic manipulator that includes a long-reach, extendable/retractable boom element, combined with an electropermanent magnetic capture head, enables capturing spacecraft at a distance, even when the two vehicles have some residual relative velocity.

The Electropermanant Magnetic Boom Assisted Rendezvous and Capture (EMBARC) technology (Figure 37) is a simple, low-cost capture solution. Since it is designed to capture objects that have already been outfitted with ferrous grappling targets, it does not require the same sophisticated force-feedback control systems that the Front-end Robotic Enabling Near-term Demonstration (FREND) arms on RSGS use to “grab-on” to non-cooperative legacy Geosynchronous Earth Orbit (GEO) communications satellites that do not already have ferrous grappling targets installed.

Figure 37: Bulldog LEO Satellite Servicing Vehicle - Shown with EMBARC boom-assisted magnetic capture system

Because most spacecraft do not use ferrous materials, Altius has developed a lightweight DogTagTM cooperative grappling fixture (shown below in Figure 38), which includes a ferrous capture surface that is coated with a durable optical coating that provides machine vision recognizable targets to simplify rendezvous and capture operations. Altius is in the process of flight qualifying its DogTag grappling fixture for use on one of the LEO mega constellations currently being developed. With a goal of flying on one of the constellation spacecraft in the 2019 timeframe, a follow-on rendezvous and capture demo using the Bulldog LEO satellite-servicing vehicle will be enabled in the early 2021 timeframe.

Figure 38: DogTagTM Cooperative Grappling Fixture

These capture technologies could be used in the propellant transportation architecture in two ways. One way would be to use the same DogTag grappling interfaces and associated Bulldog servicing vehicles (or scaled-up derivatives) as tugs to capture DogTag equipped tankers and/or cargo vehicles such as reusable second stages or large reusable lunar landers, and maneuver the captured vehicles safely to the propellant transfer facility. Another option would be to use scaled-up versions of the EMBARC Boom-Assisted Magnetic Capture robotics systems to enable Advanced Centaur-class vehicles to rendezvous directly with the propellant transfer facility, as shown below in Figure 39 and Figure 40.

Figure 39: Artist's Conception of a LEO Propellant Depot for Smallsat Launch Vehicles Shown using EMBARC boom-assisted magnetic capture technology for direct rendezvous/capture of smallsat launch vehicle upper stages

Figure 40: Illustration of a Soft-Capture Rendezvous with Tanker Tanker equipped with multiple EMBARC-style boom-assisted magnetic capture arms, rendezvousing with and soft capturing an upper stage and payload for distributed lift in-space refueling

Using relative navigation sensors and communications systems on the transfer facility, an incoming stage can be tracked. It is then be possible to provide maneuver commands to the arriving vehicle allowing it to enter a trajectory that drifts by the facility at a safe distance (10-20m) and low relative velocity (<5cm/s). Once in this state, the stage can be magnetically soft captured by one or more such EMBARC capture systems.

Force is gently applied to cancel out relative motion, and then retract, pulling the arriving vehicle close to the propellant transfer facility (within 1-2m) for subsequent refueling. One unique feature of the EMBARC capture system that aids in this type of capture operation is that as the extendable/retractable elements retract, they become more and more stiff in bending, enabling a gentle soft capture at a long distance, followed by more rigid manipulation when retracted close together.

Once capture has been made, one key remaining element is the cryogenic transfer coupling that enables connecting the propellant transfer facility to a visiting tanker to receive or deliver propellants. Because the proposed transfer vehicles will be derived from existing upper stages, Altius is developing a cryogenic transfer coupling (one version of which is shown below in Figure 41) that can serve as an upper stage fill/drain T-0 disconnect coupling. The airborne half of the coupling is being designed so that it can be easily robotically reconnected in-space for subsequent propellant transfer operations - either in zero gravity or on the lunar surface.

One of the key unique features of the Altius cryogenic coupling is the ability to “deactivate” the cryogenic seal for low-force insertion/extraction, and then “reactivate” the seal once coupled to form a leak-tight connection. This low-force insertion/extraction characteristic is important for robotic propellant transfer connections. Altius is currently developing this technology under a NASA SBIR Phase II contract, with the goal of flight qualifying a subscale LOX version of this coupling for flight demonstration on a small satellite launch vehicle in the 2019 period, with subsequent development of an upper stage scale LH2-compatible version in follow-on efforts.

Figure 41: Cryogenic Transfer Coupling Early conceptual design for a dual-use cryogenic transfer coupling with deactivatable cryogenic sealing sections

In-Door” Robotics

In addition to external robotics, the NRHO facility will almost certainly require internal robotic systems for normal operations, maintenance and repair. Two examples of such systems are shown in Figure 42. These could perform various functions including unpacking and stowage of replenishment items, inspection and repair of anomalies, and facility maintenance and cleaning.

Figure 42: Internal Robotics for Mission Support (left) an illustration of the smaller segment of Canadarm 3 inside the Gateway

NASA’s R5 aka Valkyrie

NASA’s R5 aka Valkyrie was designed and built by the Johnson Space Center (JSC) Engineering Directorate to compete in the 2013 DARPA Robotics Challenge (DRC) Trials. Valkyrie, a name taken from Norse mythology, is designed to be a robust, rugged, entirely electric humanoid robot capable of operating in degraded or damaged human-engineered environments. Building on prior experience from designing Robonaut 2, the JSC Valkyrie team designed and built this robot within a 15 month period, implementing improved electronics, actuators and sensing capability from earlier generations of JSC humanoid robots.

Following the robot’s appearance at the 2013 DRC Trials, the Valkyrie team modified and improved the robot - modifying the hands to increase reliability and durability, redesigning the ankle to improve performance and upgrading sensors for increase perception capability. The Valkyrie team also partnered with the Florida Institute for Human and Machine Cognition (IHMC) to implement their walking algorithms on NASA hardware in preparation for the Space Robotics Challenge, part of NASA’s Game Changing Development Program and Centennial Challenges.

Power/Battery

Valkyrie can be configured to run from a wall or from battery power. The custom dual-voltage battery is capable of running the robot for about an hour. When a battery is not in use, it can be replaced with a mass simulator and capacitor that simulates the mechanical and some of the electrical properties of the battery.

Head/Sensor Suite

Valkyrie’s head sits atop a 3 DOF neck. The main perceptual sensor is the Carnegie Robotics Multisense SL, with modifications to allow for IR structured light point cloud generation in addition to the laser and passive stereo methods already implemented. Valkyrie also features fore and aft “hazard cameras” located in the torso.

Arms

Each upper arm consists of 4 series elastic rotary actuators and when combined with the forearm has 7 joints. The arm has a quick mechanical and electrical disconnect between the first two joints that allows for easy shipping and service.

Forearms/Hands

Valkyrie features a simplified humanoid hand, with 3 fingers and a thumb. Each forearm consists of a single rotary actuator (realizing the wrist roll), a pair of linear actuators (realizing wrist pitch and yaw), and 6 finger and thumb actuators. The hands are attached to the ends of the arms with mechanical and electrical quick disconnects that allow for easy shipping and service.

Torso/Pelvis

The robot’s torso houses two series elastic rotary actuators (the first arm joint on either side), two series elastic linear actuators that work in concert to realize motion between the torso and pelvis, and various computer and power facilities. The pelvis houses three series elastic rotary actuators: the waist rotation joint, and the hip rotation joint of each leg. The pelvis is considered the robot’s base frame, and includes two IMU’s.

Legs

Each upper leg contains five series elastic rotary actuators. The ankle is realized using two series elastic linear actuators working in concert. The leg has a quick mechanical and electrical disconnect between the first two joints that allows for easy shipping and service.

Specifications

Weight: 300 pounds Height: 6 feet 2 inches Battery Energy: 1.8kWh Computers: 2 x Intel Core i7