Proving the Moon Before Launch

How Astroport sized its lunar fleet before committing to hardware

Astroport Space Technologies needed to prove its autonomous lunar site-preparation vehicles would move, climb, and last on the Moon before building them. SoftServe ran physics-based terramechanics simulation that validated both vehicle designs and quantified their power, torque, and energy demands across the full operating envelope — giving Astroport what it needed to size the mission’s batteries and charging infrastructure before committing any hardware.

Introduction

You can’t test a vehicle where it is going to work. That is the problem with building on the Moon.

Astroport Space Technologies is developing the infrastructure that will make a sustained lunar presence possible — beginning with the landing pads, roads, and prepared surfaces that nearly every other mission depends on. Founded in 2020 and working under NASA’s Small Business

Technology Transfer (STTR) program, the San Antonio company is designing a site-preparation architecture and concept of operations (CONOPS) in which autonomous machines excavate lunar regolith, haul it, and process it into durable construction material.

That architecture rests on two ground vehicles: a heavy hauler that moves loose regolith across the surface, and a general mobility machine that carries excavation and manipulation tools. Both must perform in an environment no terrestrial test track can reproduce — one-sixth gravity, abrasive and loosely packed regolith, and no chance of a service call once they are deployed.

Before committing to these designs, Astroport needed answers to some deceptively simple questions. Would the wheels sink? Could the vehicles climb the grades the mission requires while fully loaded? How much power would they draw — and how much battery and charging infrastructure would the whole operation have to carry to the surface to keep them running? Getting those numbers wrong on Earth is expensive. Getting them wrong on the Moon can end a mission

So Astroport brought in SoftServe to answer them — not with rules of thumb, but with the same class of physics-based simulation that has guided off-world vehicle design since Apollo.

Challenge

Designing for a surface you canʼt test on

Astroport’s site-preparation plan only works if its vehicles can move reliably across lunar regolith. That regolith is a loose, abrasive, poorly understood material — it behaves nothing like terrestrial soil and cannot be faithfully recreated on Earth at one-sixth gravity. The company had credible reference concepts for both vehicles, but it lacked the in-house terramechanics and geotechnical simulation capability to prove those concepts would actually perform.

Terramechanics did not begin in space — it grew out of terrestrial off-road, military, and agricultural vehicle engineering, and that is exactly where it pays off again. The same simulation that sized Astroport’s lunar vehicles applies directly to mining and earthmoving: predicting traction, sinkage, gradeability, and haul-cycle energy for heavy fleets on soft or graded ground, so operators can size equipment before committing capital. It extends to heavy off-road transport in sectors such as oil and gas, wherever vehicles must move reliably over difficult terrain.

The specific unknowns were the ones that decide whether a surface vehicle succeeds or strands itself:

• Wheel sinkage and rutting: How deep the wheels settle into regolith, and how that changes from an empty to a fully loaded vehicle.
• Motion resistance and drawbar pull: The usable force a vehicle generates after overcoming its own rolling resistance — whether the vehicles can move at all and, for the tool-carrying machine, whether it can push or pull implements through the soil.
• Gradeability under load: Whether a loaded vehicle can still climb the slopes the mission demands.
• Drive torque, power, and energy per distance: The “mileage” that determines how much battery and charging infrastructure the mission must provide.

These were hard to pin down: the hauler’s mass changes dramatically between empty and fully loaded, and the heritage data that exists is scattered across decades of separate programs. Without a model that ties it together under lunar conditions, every design choice was a guess.

The risk cut both ways. Oversize the vehicles and their motors, and Astroport wastes launch mass and power it cannot spare on the Moon. Undersize them, and a vehicle can bog down, fail to climb, or strand itself — compromising not just site preparation but every mission that depends on a prepared surface.

Solution

Validated terramechanics, from Apollo to Astroport

Across Astroport’s STTR phase, SoftServe’s Robotics and Advanced Automation team ran the terramechanics simulation with one clear goal: turn two notional vehicle concepts into designs backed by numbers.

The method has a long track record in off-world vehicle design. Its simulations use semi-empirical terramechanics in the Bekker-Wong-Reece tradition — a framework built on decades of off-road, military, and agricultural vehicle engineering. The same modeling tradition informed the Apollo Lunar Roving Vehicle, China's Yutu lunar rovers, the Zhurong Mars rover, and Europe's ExoMars rover. That heritage matters: the predictions rest on models calibrated against real wheel-and-soil test data and proven across decades of rover design, rather than on first-principles guesswork — with the lunar regolith parameters themselves treated as the key modeled input to be refined as ground-truth data improves.

The team built parametric models of both vehicles in the site-preparation architecture: the Astroport Transport Platform (ATP), a baseline hauler design, and Astrolab’s Flexible Logistics and Exploration rover, FLEX. “Parametric” is the key word — the models let Astroport step through ranges of vehicle geometry, wheel dimensions, wheel stiffness, and load conditions, and see immediately how each choice affects performance.

For every configuration, the simulations predicted the quantities that decide whether a design works:

• Wheel sinkage and rut depth
• Motion resistance and drawbar pull versus wheel slip
• Gradeability, or slope-climbing ability under load
• Drive torque
• Drive power and energy consumption per distance traveled
• All of it was computed under lunar gravity and the regolith-interaction parameters defined for the project, so the outputs reflect the environment the vehicles will actually face.

The team implemented the models in OpenModelica with Python, and used NVIDIA Isaac Sim to visualize the rover in its operating environment. The result was less a single answer than a design instrument: a way for Astroport to choose wheel sizing and motorization on a rational basis, confirm whether existing actuators are adequate, and — critically — translate vehicle performance into the power and charging infrastructure the mission must carry.

Value Delivered

Designs validated, sized, and proven

The headline result is the one Astroport needed most: the simulations confirmed both baseline vehicle designs are suitable for their site-preparation architecture — including energy consumption and traction. Astroport can move forward on its mobility and power architecture having de-risked it before committing any hardware.

What the modeling established:

• Power budget: A model of the hauler’s electrical demand across its full operating envelope — from steady level hauling to worst-case loaded slope climbs — gave Astroport a complete power budget confirming it stays within what the platform can supply.
• Drivetrain: The analysis confirmed that the existing actuators are correctly sized across the operating range, so the drivetrain does not need a costly redesign.
• Battery and charging: By predicting how much energy the vehicles consume as they work, the modeling let Astroport size batteries and place charging stations across the site — turning an open power question into a concrete infrastructure plan for its CONOPS.
• General mobility machine: The analysis confirmed the mobility machine generates more than enough drawbar pull to drive several excavation tools through regolith.

Each of these findings is the line between a rover that climbs the grade and one that strands itself, and between a power system Astroport can afford to launch and one it can’t. Together, the work turned two unproven concepts into a quantified design envelope — wheel sizing, motorization, traction, and power infrastructure — that Astroport can build against.