The AI Space Factory Mars Habitat construction concept of using a 25% PLA plastic + 75% Mars Basalt fiber mix is one of the more promising and innovative solutions in development. This system could potentially build rigid shelters on Mars using material made on Mars and will offer pressurization, and protection from radiation and weather. 700 x 700 pixel version.

Unveiling the science fact and fiction about AI Space Factory’s 3D printed Mars Habitat

This article conducts an in depth analysis of a very intriguing concept being developed by AI Space Factory: a 3D printing a PLA/ Basalt fiber Mars Habitat they call MARSHA.

The team here at The Mars Blueprint happens to include 3D printing enthusiasts and so this piece is going to dig into the AI Space Factory MARSHA concept (short for MARS HABITAT) and the feasibility of it working on Mars for in situ habitat production. So be warned, some serious geeking out is about to occur!

For the TLDR crowd…. The AI Space Factory concept of using a 25% PLA plastic + 75% Mars Basalt fiber mix is one of the more promising and innovative solutions we have seen. This system could potentially build rigid shelters on Mars using material made on Mars and will offer pressurization, and protection from radiation and weather. At this point the project is equal measure science fact and fiction. The 3D printing tech is completely science fact but the access to materials on Mars is pretty much fiction – for now. To make the habitats feasible the initial builds will most likely need to be smaller than the proposed 45 ft x 24 ft design and some major engineering will need to be done to produce the PLA and basalt fibers on Mars. This last part is the major design show stopper with what is publicly known at the present. There is not a fully developed plan (that we know of) for large scale production of PLA and basalt fiber on Mars. Some of these technical challenges can be overcome by sending materials for the initial habitats from Earth, but the large volumes required for a full-scale habitat make that prohibitive. Extensive development into automated crop production and/or yeast bio reactors for PLA production and an automated foundry and fiber spinning production facility is needed to make this concept a viable part of a plan for Mars habitats.

For those still wanting to take the technical plunge, buckle up because as Mark Watney once said… it is time to science the shit out of this. On to the deep dive analysis!

To baseline everyone, in 2019 AI Space Factory won the NASA and Bradley University sponsored 3D Printed Habitat Challenge to 3D print a 1/3 scale Mars Habitat in 30 hours or less. The two finalists in this competition were AI Space Factory and Penn State University. Penn State University used a more tried and true concept that is already in use here on Earth printing a quick drying concrete mixture. AI Space Factory used a much more innovative concept that mixed 25% Polylactic Acid (PLA) plastic with 75% basalt fibers which they believe can be made from Martian sourced basalt.

The reason the AI Space Factory concept is so intriguing is that PLA is one of the most user friendly plastics on the market and is widely used with Fused Deposit Modeling (FDM) 3D printers. This type of printer lays down one layer of molten plastic at a time to gradually build a complex shape like the MARSHA habitat. PLA is known as a bio polymer as it is derived from organic material such as the starches from corn or other plants. It can also be produced by genetically engineered yeasts.

AI Space Factory 3D Printer: 100% Science Fact

We will start our deep dive by looking at the AI Space Factory 3D printer. In the imagery it looks like AI Space Factory repurposed a multi-axis robotic arm that is commonly seen welding in a factory setting. As this is designed to function (someday) on Mars, power constraints are going to be a fact of life and need to be thoroughly planned out before it ever leaves Earth. From an energy standpoint, this type of industrial robot uses somewhere in the ballpark of 3 KW/ hour power consumption. This is within the 35-40KW range NASA estimates a base will need and for perspective is also roughly the median daily power usage for an average American home here on Earth. To generate 40KW with Solar panels it would require roughly 4500 square feet of panels on Mars. This is based on a rate of 10 KW per 525 square feet on Earth and adjusting for Mars being further from Earth and only getting 43% of the sunlight.

In the imagery and videos we can see that AI Space Factory switched out the welding tool head with a custom large diameter extruder/hotend/nozzle print head. In looking at the pictures and videos it appears the setup uses a custom direct drive extruder heater combo that goes down the central core through the cooling disk. Given that PLA prints at ~200 degrees Celsius, the design appears to be focused on ensuring the large diameter hotend provides a consistent PLA flow despite the 75% fiber mix. The interesting part of this setup is that the system appears to use a custom part cooling blower setup that uses 12 cooling tubes connected to high volume blowers mounted lower on the robot arm. This part was not seen before the competition but is commonly added to 3D printers to allow for the plastic to cool more rapidly which decreases warping and ensure the structure is cooled to a more rigid state faster for stability.

To estimate the power consumption of the hot end we will start by guestimating that hot end has a 4cm nozzle based on the images of the nozzle and print layers. In our research, resistive heaters like those used in a typical 3D printer have a linear increase in power consumption that corresponds with their size. This was used to estimate a 4cm nozzle would draw roughly 100x the power that a .4 mm nozzle uses (average 70 – 100 W/hour) which means the large hotend would require around 7 – 10 KW per hour. We cross checked our math by using an open source equation for calculating wattage in a heating element (Watts = 0.018(Aluminum coefficient) x Lbs of heating element x ΔT (~400 degrees F/200 degrees C) / Heat-Up Time (in hrs)). Combined with the 3 KW requirements for the robotic arm the setup would require 10-13 KW/hour which would be readily available from the solar panels during daylight hours.

The NASA competition showed that the printer is completely capable of printing a shelter in an environment with constrained weather and temperature. The one variable that can’t be constrained on Mars is the extreme temperatures which will have a negative impact on the PLA. According to NASA the average temperature is -81 degrees F (-63 degrees C). A summer day on the equator can swing between 70 degrees F (20 degrees C) to negative 100 degrees F (negative 73 degrees C). Dropping below 50 degrees F (10 degrees Celsius) is known in 3D printing circles to cause PLA layers to lose adhesion to each other and/or start warping or curling as the plastic cools too rapidly. While the basalt fibers may mitigate some of this as they act similar to CF infused PLA and bind together much better than normal PLA, it will likely need some sort of software solution to pause the print at night when temperature plummets and power is limited.

PLA Production: 50% Fact / 50% Fiction

PLA as a printable building material is 100% science fact. It is by far one of the most widely used 3D printing materials on the market and the AI Space Factory concept is just an extreme example of how flexible it is. PLA is also very easy to produce as it technically can be made out of any organic process that produces lactic acid which is then converted into Polylactic Acid. Companies like Naturemade currently use readily available starch from plants such as corn, which they then have to process into lactic acid that can be turned into PLA. If you read our “Beer, Yogurt and 3D Printers… The future of Mars cuisine” article you would see there is some amazing work being done today in the fields of engineering yeast to produces flavors, proteins and many other needed components of food and medicine. Naturemade and other companies are looking into taking it a step further to directly produce PLA. For instance, the yeast Yarrowia Lipolytica has been modified to directly produce PLA. An entire article could be written on the processes and requirements to run a commercial level bioreactor to produce PLA on Mars, so we are simply going to say that the baseline science is already there. The part that is 50% science fiction is that the technology needs to be matured to make PLA on Mars in large quantities. The same technology needed to use yeast to produce food can also produce PLA albeit with some serious engineering and automation needed to produce the tons of material needed for a habitat.

Basalt Fiber Production: 50% Fact / 50% Fiction

The second component of AI Space Factory’s blend is basalt fiber. Mars surface is covered in volcanically produced Basalt and Basalt fiber is made by melting basalt and using a process called pultrusion to produce the fiber. This part is completely fact and this amazing material is used all around the world right now. Basalt fibers have similar properties to fiberglass or carbon fiber especially when mixed in with the PLA giving it much higher strength, layer adhesion and resistance to damage. As with the production of PLA, the part that is largely science fiction is in producing basalt fiber on Mars. There is no real equipment designed to do this on Mars- at the present – which therefore causes a host of engineering challenges likely to cause a significantly delay in the implementation of this type of structure until a fairly large manned footprint has been established. The reason for this is that to produce basalt fibers it is required to collect tons of basalt, crush it, heat it to 1350 degrees Celsius and then carefully pull it onto a specialized spindle to create the fiber. These are all labor and energy intensive processes that currently only exist on Earth. Based on the size of the proposed habitats nearly a hundred tons of fiber would be needed for each full-scale habitat. The energy requirements alone make this out of reach for an early mission as a modern electric kiln draws 2500+ KW per hour to produce 5-10 tons of molten ore which far exceeding the power generation capabilities of an early base.

Transporting raw materials: 75% Science Fiction

Producing PLA and Basalt fibers on Mars will take time so we are going to assume that the first few habitats are printed using materials brought over on the first lander. As any enthusiast in the 3D printing community can tell you, the rolls of filament disappear at a much higher rate when you scale up a print. Printing a Mars habitat is no exception. To run the calculations on the amount of PLA the AI Space Factory system would need to build their proposed 45 foot tall x 24 foot wide habitat we used Cura 5.1 and built a custom profile to match the build. As all of the nozzles and dimensions for printers are metric, the rest of this analysis will stick to metric units for simplicity.

AI Space Factory’s PLA/Basalt fiber mix is proprietary, so the exact weight of the mix in comparison to pure PLA is unknown. This is a key detail that will impact whether the first habitats can have the raw materials shipped over from Earth or if it is too heavy and will have to be completely produced on Mars. To get a ball park figure, we did a little math based on material density. Basalt fiber (1.796 g/cm3) is 30% more dense than PLA (1.24 g/cm3). Based on this we project the 75% fiber component of the mix adds approximately 30% more weight than its PLA equivalent and therefore we adjusted our weight guestimates accordingly.

The metric equivalent of a full-scale MARSHA is 13.716 meters tall by 7.313 meters wide. Based on imagery of the print and nozzle at the competition we guestimated a 4cm nozzle which simplifies later assumptions as this is a 100x scale up from a standard hobby sized .4 cm nozzle. Based on the imagery and videos we set Cura to make the structure two walls thick. At this scale, Cura estimates that a full scale 100% PLA structure with an assumed interior material usage of 20% infill to account for stairs, supports and basic amenities would require 127.998 metric tons of PLA. Scaling that back to 25% as per AI Space Factory’s stated mix means that a full scale habitat would require 31.9995 metric tons of PLA and accounting for the density offset, 124.798 metric tons of basalt fiber. The PLA alone is approximately half of the current stated launch ability of the SpaceX Starship which has a theoretical max lift capacity of 90+ metric tons. The combined PLA + basalt fiber would be roughly 156.8 metric tons and would take nearly two Starships to just transport the raw material for one habitat. Therefore at full scale, this does not appear to be science fiction as it exceeds the stated Starship lift ability.


As any 3D printing enthusiast can tell you, increasing print scale tends to increase the material usage at an exponential versus a linear rate. In the case of MARSHA this is because it is basically a curved cylinder. The equation for a cylinder’s volume is V = π (R2 -r2)h where R is the outer radius and r is the inner radius. As the radius gets larger the printed material volume increases exponentially. This means that the opposite is also true and scaling back the size of a print also gives massive advantages for weight savings. Dropping the habitat to a 2/3 size requires 41.04 metric tons at 100% PLA and 10.26 metric tons of PLA at 25%. Basalt fiber would add another 40 metric tons for a total of 50.26 metric tons of raw materials. The one third scale model that AI Space Factory demonstrated at the competition has the greatest savings requiring only 6.36 metric tons at 100% and 1.59 metric tons of PLA at 25%. Basalt fiber would add another 6.2 metric tons for a total of 7.79 metric tons. To bring this concept over from fiction to fact, AI Space Factory would need to make their initial MARSHA on Mars closer to the 50-70% scale. This would ensure the weight of the PLA, basalt fibers, and printing equipment is within the lift capability of the Starship as the lower weight would take up around 50% of the total weight capacity.

Thank you to those who made it this far and came along on our brief descent into 3D printing madness. Bottomline… there is a lot of work to do here, but this concept shows immense promise once the engineering work is done to produce PLA and Basalt Fibers on Mars. We firmly believe AI Space Factory is on the right path but there are a lot of technical challenges involving making or transporting the PLA and Basalt Fibers Mars required to make this a viable option at the present. However, this concept meets a key design principle of the Mars Blueprint in ensuring the future missions are able to “live off the land” and maximize in situ production of raw materials whenever possible.

We had a blast running the numbers, but we are by no means infallible. So please let us know if you think we missed something here or are wrong on our conclusions.