The Lunar Geotechnical Manifesto

A frank assessment of where the space industry stands, and what it continues to overlook

Lunar activity is accelerating. NASA is preparing for sustained surface operations under Artemis; SpaceX and Blue Origin are designing vehicles capable of delivering unprecedented mass to the Moon; and a growing cohort of robotics start-ups is developing excavators, haulers, drilling systems, and in-situ manufacturing technologies. Each organization plays a necessary role in shaping the emerging lunar economy.

Yet all of this momentum shares a structural omission:

none of these systems are being designed against an engineering definition of the lunar ground.

For decades, lunar exploration has been guided by science missions, mobility experiments, and technology demonstrations. These have built an extraordinary foundation for understanding the Moon as a planetary body. But they do not provide what civil engineering requires for construction: the mechanical behavior of the material that must support reactors, pads, roads, trenches, storage modules, and structures produced from regolith.

Infrastructure demands a different type of knowledge, subsurface knowledge, and today it remains largely absent.

This is not a controversial statement. It is a technical one.

There is still no lunar or analog ASTM standard, no construction code, and no geotechnical guideline derived from in-situ subsurface data. Instead, today’s design assumptions depend on:

• literature reviews,

• laboratory simulants developed for traction and dust behavior,

• reconstructed Apollo-era anecdotes,

• and surface imagery interpreted through scientific, not engineering, lenses.

These inputs are valuable for research, but insufficient for predicting excavation response, bearing capacity, settlement, compaction pathways, and long-term ground performance, the very elements on which lunar infrastructure depends.

NASA’s architecture, SpaceX’s lander mass envelopes, Blue Origin’s surface-cargo concepts, and the robotics roadmaps of multiple start-ups all presume some level of regolith cooperation. But cooperation cannot be assumed. It must be engineered.

The Problem Behind the Progress

Across the United States, China, India, Japan, Europe, and the private sector, lunar surface systems are now being designed with confidence: reactors that require stable foundations, landers delivering increasing mass, regolith-based construction concepts, autonomous fleets for transport, subsurface storage, and ISRU systems intended to operate continuously for months or years.

On paper, this looks like rapid advancement.

In practice, these plans are being built on an undefined engineering substrate.

None of these architectures, whether government-led or commercial, are grounded in a construction-grade understanding of how lunar regolith behaves under load, excavation, or prolonged operation. The topic circulates in private meetings and side discussions, yet it rarely appears in published design documentation or roadmap material. The absence is not accidental; it reflects a deeper assumption that the ground will be cooperative enough, or that the uncertainties can be absorbed by oversizing equipment and increasing power budgets.

But the evidence does not support that assumption.

From Apollo’s early penetration measurements to Chang’e’s sampling data and the most recent CLPS results, the same pattern recurs: the regolith is structured, mechanically resistant, and far from homogeneous. Shallow refusal depths, density inversions, and pre-loading effects have appeared at different sites, across different missions, and under different operational conditions. These are not anomalies. They are signatures of a material whose stress history and internal structure have never been modeled for construction.

These observations should have prompted a shift toward formal geotechnical modeling years ago. Instead, the sector continued forward with spacecraft-driven priorities, relying on approximations rather than mechanical parameters. As a result, the most expensive elements of future missions, excavation systems, surface power units, and regolith-handling machinery, are being designed to operate on ground conditions that have not yet been characterized in engineering terms.

Mobility Testbeds Are Not Construction Testbeds

Most of today’s lunar testbeds, across government labs, universities, and private facilities, are built around mobility: traction studies, wheel-soil interaction, slip ratios, dust lofting, obstacle negotiation, and autonomous navigation. These are valuable activities, but they do not address the engineering questions that determine whether infrastructure can be built or maintained.

None of the current testbeds simulate drilling behavior in high-OCR* regolith, compaction response in low-OCR* material, or excavation resistance at depth. Very few attempt to quantify tool wear, cutting energy, or realistic bucket mechanics. Innovation in drilling systems for structured regolith remains limited, and the industry still treats penetration resistance as an operational inconvenience rather than an engineering parameter.

Gravimeters, which could reveal density variations and subsurface layering, are still underutilized. Ground-penetrating radar, widely proposed as the default payload for subsurface investigation, is not a substitute for mechanical testing; it identifies interfaces, not how the material behaves under stress.

This is the gap that matters.

Construction on the Moon requires answers to questions that mobility studies cannot provide:

• How much energy does excavation require at depth?

• How does the regolith compact under repeated loading?

• What is the bearing capacity for reactors, radiators, or tanks?

• How will pads deform under thermal cycling and operations?

• Which regions are actually buildable without major ground improvements?

These are not academic questions; they are the prerequisites for any serious surface architecture. Yet the tools normally used to answer them, Geotechnical Data Reports (GDR), Geotechnical Interpretation Reports (GIR), Geotechnical Baseline Reports (GBR), classification systems, and in-situ testing protocols, simply do not exist for the Moon.

Mobility research can tell us how to traverse the surface.

It cannot tell us where we can build, how we can excavate, or whether the ground can support the infrastructure we intend to deploy.

The Misconception of "Considering Geotechnical Properties"

The phrase appears often in agency, academic, and commercial documents: “geotechnical properties have been considered.”

But the claim usually collapses under inspection.

Most organizations still rely on The Lunar Sourcebook and a handful of legacy mission reports. These works are essential scientific references, but they were never designed to support construction engineering. They describe the Moon; they do not prescribe how to build on it. The industry has remained in that descriptive phase far longer than necessary.

When teams say they have “considered geotechnical properties,” they are typically referring to bulk density, particle-size distributions, cohesion values taken from simulants, or thermal conductivity. These are scientific descriptors, not engineering parameters. They offer almost no insight into excavability, compaction response, bearing capacity, settlement, deformation, or long-term structural stability.

Civil engineering operates on entirely different foundations: stress history, internal structure, variability, and load response. None of these are currently defined for the Moon in a manner that supports real design decisions. As a result, construction concepts drift away from mechanical reality and performance predictions become speculative rather than engineered.

What is missing is not just data, it is planning.

Even in the absence of complete subsurface datasets, lunar equivalents of a Geotechnical Data Report (GDR), Geotechnical Interpretation Report (GIR), and the future Geotechnical Baseline Report (GBR) should already be in development. These documents structure how ground information is collected, interpreted, and used, long before the final values are known. They prevent design drift, reduce uncertainty, and keep infrastructure discussions disciplined rather than dominated by academic extrapolation or speculative science.

Without GDR, GIR, and eventual GBR foundations, conversations about lunar construction risk becoming untethered from the engineering principles that govern every terrestrial project. The industry cannot afford that divergence as it moves toward reactors, pads, ISRU plants, and regolith-based structures.

Until the community commits to treating the regolith as an engineering material, with frameworks such as OCR*, LRC, and CSI, statements about “considering geotechnical properties” will remain largely nominal rather than actionable.

The Frameworks That Are Missing

In terrestrial engineering, ground uncertainty is controlled through a structured progression: first collecting data, then interpreting it, and finally establishing a baseline that defines how the ground is expected to behave during construction. This sequence, data, interpretation, baseline, is what allows complex infrastructure to be designed, priced, and executed without ambiguity.

Lunar programs have not yet adopted this approach. There is no equivalent set of documents that describes what is known, what is assumed, or what is contractually binding regarding the behavior of the regolith. In the absence of this structure, every surface concept rests on assumptions rather than an engineering foundation. It is the single most significant gap in lunar construction planning, and it quietly affects every downstream design decision.

My work in lunar geotechnics, including the development of OCR* (Lunar Overconsolidation Ratio), LRC (Lunar Regolith Classification), and the Construction Suitability Index (CSI), was created to address precisely this problem. These frameworks translate the Moon’s surface from a scientific description into an engineering material with parameters relevant to excavation, load-bearing capacity, and site selection. They are extensions of standard civil engineering practice, adapted for the realities of planetary environments, and they provide a path toward defining which areas can support infrastructure and which cannot.

What the industry lacks is not ambition, but the structured methodology that makes construction feasible. Until that gap is closed, lunar infrastructure will continue to be designed with uncertainty that no terrestrial project would accept.

The Economic Consequences of Ignoring Ground Mechanics

In any capital project, uncertainty in ground conditions is one of the most expensive variables. On the Moon, that uncertainty does not merely increase cost, it reshapes the entire economic profile of a mission. Without a defined mechanical model of regolith behavior, engineering teams are forced into conservative design assumptions that cascade into mass penalties, inflated power budgets, and oversized systems. These choices are not technical preferences; they translate directly into higher launch costs, reduced payload capacity, and diminished operational flexibility.

A system designed for an unknown subsurface must be built to survive the worst case. That drives up CAPEX. Once deployed, it consumes more energy, cycles hardware faster, and limits duty time, driving up OPEX. The result is a surface architecture that is financially burdened before it even begins operations.

Tool wear accelerates when material structure is misunderstood. Excavation loads deviate sharply from predictions based on simulants. Foundations deform if compaction and settlement characteristics are assumed rather than measured. On Earth, these issues lead to cost overruns and delays. On the Moon, they can force mission redesigns, relocation of high-value assets, or even the abandonment of planned infrastructure zones.

The economic risk compounds when infrastructure is placed on terrain that cannot support its functional load. A reactor installed on a pad with unverified bearing capacity is not merely a technical hazard, it is a stranded investment. An ISRU plant located in a zone with poor thermal cycling behavior is not simply inefficient, it becomes an ongoing operational liability. A road or logistics corridor built across mechanically weak regolith does not just fail structurally; it constrains every downstream activity that depends on it.

The financial sector understands this intuitively. Uncertainty is the most expensive line item in any project. The lack of geotechnical definition on the Moon forces every contractor, integrator, and robotics provider to build unnecessary margin into mass, torque, power, tooling, and redundancy. That margin accumulates across the entire architecture. It is paid for multiple times: in launch cost, in energy, in hardware fatigue, and in reduced operational envelope.

Nearly all of this inefficiency is avoidable. If the regolith’s mechanical behavior were defined through structured geotechnical investigation, the lunar equivalents of GDR, GIR, and future GBR, system mass and power requirements could be optimized rather than inflated. Buildable zones could be identified early, avoiding the costly mistake of placing infrastructure where it cannot remain stable. Excavation and construction equipment could be sized for the real operating environment rather than hypothetical extremes. And the economic model of lunar operations would shift from risk-absorption to predictable performance.

In short, geotechnics is not a scientific luxury, it is a financial necessity. The cost of ignoring the ground will appear not in academic margins, but in every CAPEX and OPEX figure associated with lunar construction. As ambitions move toward reactors, habitats, industrial pads, and ISRU systems, the absence of subsurface knowledge has become the single most under-recognized economic risk in the entire lunar program landscape.

A Direct Message to the United States Government and Artemis Leadership

The United States retains unmatched capability in launch systems, deep-space operations, spacecraft autonomy, and scientific analysis. Artemis is a technically ambitious program backed by decades of institutional experience.

But none of this guarantees the U.S. an advantage where it now matters: the ability to build and operate reliably on the lunar surface.

China is executing a phased lunar program that deliberately accumulates the operational knowledge required for construction. India is advancing with targeted missions that close critical gaps in polar operations and regolith characterization. Both are integrating engineering considerations into mission planning at an earlier stage than the United States.

The U.S. approach still treats the mechanical behavior of the lunar regolith as a peripheral issue. In practice, it is a primary design constraint. NASA, its contractors, and commercial partners are preparing to deploy reactors, landing pads, mobility systems, regolith-handling equipment, and ISRU plants without a construction-grade subsurface model. Nothing in the current Artemis documentation provides the engineering definition required to size excavation systems, design foundations, or determine where infrastructure can be placed without unacceptable deformation or settlement.

Apollo, Chang’e, and CLPS missions have already demonstrated shallow refusal, structural layering, and pre-loading effects across widely separated sites. These behaviors will affect every excavation tool, every pad, every structure built from regolith, every buried component, and every system requiring long-term stability. Yet these observations have not been translated into parameters usable for design.

U.S. leadership should be aware that China is quietly incorporating this reality into its program sequencing. India is moving in the same direction, building capacity through incremental, targeted measurements. The absence of an American geotechnical framework is no longer an academic gap, it is a strategic one.

If Artemis is intended to be more than a series of landings, the program must take the next step:

define the ground in engineering terms.

This requires producing the lunar equivalents of a Geotechnical Data Report, a Geotechnical Interpretation Report, and ultimately a Geotechnical Baseline Report. It requires integrating regolith mechanics into system design before hardware is finalized. And it requires adopting or developing frameworks that translate scientific observations into engineering decisions.

This is a standard practice in every major terrestrial infrastructure program. Without it, costs rise, systems are oversized, operational timelines degrade, and the risk of placing critical assets in unsuitable terrain increases significantly.

The reality is straightforward:

The United States is preparing to construct on a surface it has not yet defined. Competitors are moving to close that gap.

Addressing this does not require new rhetoric or new slogans, it requires the same engineering discipline the U.S. applies to tunnels, reactors, foundations, and long-duration structures on Earth. Nothing more, and nothing less.

The United States has the capability to lead on the Moon, but capability is not leadership. Leadership begins when a program defines the conditions it intends to build upon. Right now, the regolith remains undefined in engineering terms, and every surface architecture, American, Chinese, Indian, or commercial, rests on that omission.

If Artemis is to succeed as a construction program, not just a landing program, the ground must become a design input, not an assumption. Until then, the most advanced hardware in the world will be operating on uncertainty, while others are learning to operate on knowledge.

The nation that characterizes the lunar ground first will set the terms for everyone who follows.

Roberto Moraes

Author