Building Roads and Landing Pads on the Moon

Doug Plata describes how a few tele-operated robots can make roads and landings pads on the moon.

The road would be built by robots to remove boulders. These would be followed by bulldozer bots to flatten the path. There would then be heavy roller bots to flatten and compact the regolith. A laser-equipped bot would sinter the top layer and there would be a final pass of the heavy roller bots.

The dust-free landing pads could be made with light bulldozer bots that would also layout tarps. There would lighter tarps away from the exhaust and blast-resistant tarps under the lander.

54 thoughts on “Building Roads and Landing Pads on the Moon”

  1. I have thought about using sintering to make large printed structures.

    Basically the surface would be cleared of any rocks and soil down to a level firm enough to work as a foundation. The it would be heated with solar, laser, microwave power to melt it. Then a thine layer find soil is layered on top and melt again. Repeat this may times and you would have a suitable road or landing pad or foundation.

    To make a structure with rooms repeat the lay and melting process But leave the place were the room it to be cold so it doesn’t melt or sinter. This would require a printer type sintering or melting system. Repeat as many times as necessary to make a 3D structure.

    Then finally shovel out the loose unheated material and install prefabricated metal structures and airlocks in the rooms.

    This sintering or melting printing would not need the delivery of glues, epoxies, binders from earth to make a liquid or paste type printing material. Water or other minded materials would not be needed. This would reduce the cost over any other printing method often discussed. It would however take more time to heat the material and then allow it to cool before the next layer can bee made. Power requirements are also higher but solar power is abundant for about 15 days for each luner orbit.

  2. It was a very poorly solved problem back in the Apollo days, with several cases where there were significant debris strikes one case where they pranged an engine bell on a rock, and they wound up landing Apollo 15 with a tilt >12º.

    And that was all with a lander designed to carry a payload (the ascent stage of only 4.7 t, using a 45 kN engine. If you want to get to the point where you’re safely landing 150 t, things are gonna get interesting.

    Beyond that, if something like Starship needs a rough surface landing capability to, say, put up solar curtains on the lip of a polar crater, that’s gonna create some interesting tilt requirements.

  3. No need for any of that fancy stuff. The particles are very jagged rather than smooth. This allows them once compacted with a lot of force to be very strong and durable.
    “Steam rollers” are no substitute for real high pressure brick compaction. Especially so, as you are going have to put one heck of a lot of mass on the roller because of the lower gravity.
    As there is no water involved, I suspect you want something in excess of 9,000 pounds per square inch. Perhaps much more. But whatever is needed will be cheaper than adding additives like water. Though additives like iron ore that might be close-by might be easy and cheap to add.
    Also, we are just making roads. When you are building and need to be air tight, well insulated and such, you are going to use better brick or extruded molten materials.
    If the compacting does not work (though I think this has already been demonstrated at least on Mars dust), we can always just cut rock blocks and use them to pave. No shortage of rock up there.

  4. You don’t have to maintain constant thrust. Small ballistic hops would do. You’d only activate the engines at the bottoms. Don’t need to touch the ground either.

    Though not sure if that actually saves any fuel. The stop+accelerate phase is short, but needs more thrust than a constant hover..

  5. Even fighting only 1/6 gravity you’ll burn the weight of your vehicle in propellant every ~15 minutes.

  6. The most likely location for initial human settlement is at the poles where there are both the peaks of (near) eternal light and the volatile containing cold traps in deep craters, connecting roads would be necessary, and unlike the lunar lowlands where Apollo landed the terrain is far from flat.

  7. but do you need to? Most mining operations operate without “roads” in the sense that the Moon has very compacted surfaces anyway and should easily handle large 4×4 vehicles adapted to it. Apollo rover?
    I would suspect some short distance “roads” from launch platforms to habitat is needed, but not like long highways. And if you are bring excavators and dozers to the moon, you can probably bring prefab road sections, or make them locally.

  8. If you’re making bricks, you can add water or sulfur to make lunacrete bricks, or some other binder to make a composite ceramic. Bricks are made in enclosed molds anyway, so that solves the lunacrete casting problem.

    While you’re at it, you may as well sinter them for extra strength, and maybe mix in some lunar glass or basalt fibers for even more reinforcement.

  9. It’s 1.3 lightseconds away. That’s plenty close enough. You don’t need real-time for teleoperation. Even the Mars rovers are remotely operated to an extent, and those are 20 lightminutes away (100 times longer delay).

    Fully-autonomous construction and maintenance equipment in an unstructured environment is a lot more complicated and more difficult than fully self-driving cars, and we don’t even have that yet. Partial autonomy is the way to go in the near term.

  10. There’s been a couple unmanned landings on the Moon since the 1960s, but mostly orbiters:

    I think we have a good enough understanding on Lunar soil that landing on unprepared soil is doable.

    The three versions of Spaceship that SpaceX are talking about all use the same external structure, engines, avionics, etc. The differences are only inside the payload bay. What you’re proposing is an entirely different space craft – the equivalent of designing and building a whole new house vs just changing the furniture.

  11. Problem is the humans have stopped doing it so they forgot how to do it. It was also expensive to develop back in the days.
    It is a rather big constraint for a universal space transportation system (Starship) to only be able to land on prepared infrastructure. They are already talking about three versions for different purposes. Just add a fourth optimized for vacuum only and low gravity. Take away most engines and aerodynamic features, replace the nose with docking ports and a propulsion module for horizontal landing. Move the legs etc.
    Maybe it can land vertically and then do a controlled 90 degree tip over with thusters to horizontal position. Wait… they could do that with the boosters too and end those near death experience touch-downs on the barges.

    Or just copy the Space 1999 Eagle and scale it up. Not a bad concept with modular containers for different missions.
    (And don’t forget the laser gun. There seem to be constant and endless use cases for laser guns in space)

  12. Forget teleoperating because the moon is too far away. if you have self-driving cars then you can have self-operated machines.

  13. Well the initial landing does need to deal with unprepared ground. But humans have been landing stuff on Luna since the 1960s. I think that is a solved problem.

  14. I always imagined the regolith would be sintered for a landing pad, roads to buildings and buildings themselves. I never considered that putting down a tarp would be a viable option.

  15. At our camp in Maine, we have a boulder sitting next to the shed that’s twice as high as the shed and about 5x its volume. When we tried to run a line from a new well to the camp, we discovered that we have one about the same size buried in our driveway.

    The good news about regolith is that it’s never going to aggregate, so if you can get below the center of gravity of a boulder, the force to pull it out of the fines should be pretty well quantified. But that’s a much, much more sophisticated digging and regolith-moving robot than what’s described above.

  16. The clay is terrible mostly because big globs of it sticks to your shovels and picks. And you can’t just wipe it off. You have to use your fingers. And every 3rd stroke of the pick, you are stuck and need to pry it out.

  17. San Diego can be a pain also. We have round rocks and clay after you get a foot down. The formation is called the Stadium Conglomerate. The clay holds the rocks together like a very viscus glue. The rocks very in size by location. And not everywhere has the nasty clay. In some places the matrix is red, yellow, or white. The rocks I had to dig through were about 4 in diameter, and the clay was dreadful. And it is not just a few rocks it is as many as can be packed in the space with the clay filling in the rest. The clay has grit in it that is ideal for cutting up your hands. I had to dig a 4 foot deep trench about 30 feet to lay an electrical line for the garage after paperwork was lost from when my Dad had it built. The city claimed there was no final sign-off on the construction15 years later, and with new regs the wires had to be taken down and buried. Had to do it all by hand, because the house was over 100 years old and there was no way to know where the gas lines were exactly.
    At another house the rocks are all 8 in to 2 feet still oval in yellow hardened crud. The rocks themselves when broken look white with a little purple hue in the center. And they are pretty tough to break…but break them you must…if you want to plant a pole for a fence.
    The Friars formation is the real disaster one here. Lots of slumping destroys your foundation. If a tree coming out of the ground is curved at the bottom like it started horizontal. It didn’t. The ground moved under it.

  18. Better to just collect regolith and really compact it into bricks. I think this concept has already been tested. Sure, you still need to clear and level. Then machines can place the bricks. We already have machines that place bricks on Earth roadways.
    The idea that you are going to avoid tire tracks is silly. It is like the beach. So many footprints in the sand that you don’t even notice them anymore. If it bothers people, they will rake it like they do in Japanese gardens. Though you could go neurotic and have a centrifugal dust thrower erase everything.

  19. Unless you plan to build an elevated road to get that level surface you’re going to be bulldoze, and such an elevated road uses a hell of a lot of expensive material.

  20. Instead of sintering as a few have suggested, why not apply a binding agent to solidify the road surface. Think there has been some work done already on materials which could be combined with lunar regolith to form brick like objects and a form of concrete, so this would be extending this application.

    Think someone noted as well that the dusty layer was only shallow, so if so, surely easier to simply bulldoze this to the sides and then remove rocks and fill dips. Autonomous solar powered robots could be programmed to do this, given routes, and left to the job.

  21. Right on Brian, it seems that you have facilitated a meeting of minds where your “sintered” comment, which DougSpace clarifies is not entirely correct, gets developed below by Brett Bellmore and Jean Baptiste to suggest using the abundant solar energy on the moon to melt the surface into navigable roadway alignments and building and rocket pads.

  22. But then infrastructure must be deployed on the surface with some other landing system first. And there is no guarantee one will find a good landing spot where the precursor lands. Then it must be mobile to explore multiple sites.
    I think the obvious solution is to have a lander optimized for low gravity, vacuum and unprepared ground. Some horizontal design with big feet, no heavy aerodynamic features and engine power dimensioned properly. This must be totally superior in the long run. Less hassle, more payload, less fuel use and easier cargo handling.
    Maybe hydrolox would also be better from a logistics perspective on the moon.
    Further on, we also have landings on asteroids. This would also favor a vacuum optimized, horizontal design, fueled by hydrolox.

  23. The lunar module weighed 15 metric tons, or abut 2.5 tons on the moon. I suppose one could easily build pre-fab landing pads that are properly anchored, robotically/hydraulically leveled and stable enough to handle 100 ton objects and not kick up dust. Probably not much “ground prep” required. Also, why “bulldoze”, when one could build roads with pre-fab road sections that are electromagnetically charged to keep dust at bay.

  24. I quite like the giant rocket proof tarp idea. It seems an efficient way to transport an appropriate surface with maximum area to weight.

  25. Thanks for the clarification Doug. But the dialog in your video really doesn’t explain what the laser is for.

    “Then you have like another telerobot that goes through, with like a laser that gets it really very level.”

    With the video showing the robot scanning over the whole ground surface.

    Now that you’ve explained things it makes sense. But I don’t think you can say

    “Please actually watch the animation to understand.”

  26. I can tell he’s never tried to dig a trench in Maine. It’s not the boulders on the surface that get you; it’s the ones that are buried just below the surface.

  27. I’ve mentioned solar powered sintering robots using fresnel lenses before. It would need area to contain dust and leave the sintered area behind it. Lots of bots would be needed because it would be a slow process. One problem that may be encountered while trying to rake the moon is to much dust might be kicked up in the process to block out the sunlight for solar power. Dust might settle on the panels unless they are lunar regolith-phobic.

  28. You don’t need roads on moon… there’s so little gravity that you can fly around space1999 style….

  29. What you are describing sounds like a laser scanner to create a 3D point cloud of the road for analysis and comparison to design lines & grades. However, without benchmarks (site control), the laser scan is just sort of floating in space, not tied to anything, so its use is limited. Also, a MOVING laser scan (as opposed to one fixed to a spot of known coordinates) is even worse. I guess you could use inertial methods to estimate changes in position and elevation. Maybe that’s good enough. On earth, we just use GPS controlled dozers & graders to achieve the lines & grades of earthwork design. The same could be achieved on the lunar surface with local ground-based stations and a system of benchmarks. The downside here is someone / something needs to survey the proposed route and set the benchmarks. Or just put a few GPS (LPS?) satellites in lunar orbit. BTW, the actual road construction will need to be done in several lifts, each lift compacted prior to the next. Compaction testing would be nice. Hope this is of some use. I really like the concept.

  30. Yeah, I never advocated the use of sintering for the production of landing pads. Rather, as the audio in the animation indicated, the approach would use tarp material to prevent the exhaust from interacting with the regolith and sandblasting structures.

  31. I never advocated the use of laser sintering. Rather, I advocated only the compacting of lunar regolith for the Level 1 dirt roads using electric “steamrollers”. Level 2 dirt roads could be sintered but that would best be done with microwaves and it would require a tremendous amount of energy for the length of roads that we’re talking about.

  32. The ground would have to give way by a fairly large amount in order to place the center of gravity beyond the legs. Below the immediate surface (about four inches) the regolith is pretty compacted. This was a problem during the Apollo program when they tried to take core samples.

  33. Hi Brett. I never advocated using lasers to melt the lunar regolith. Please visit for an explanation.

  34. I am the producer of this animation. Unfortunately the laser thing is generating a lot of confusion. Please actually watch the animation to understand. You can also read the description at:

    The laser is NOT in order to sinter the regolith (lunar dirt). Rather, it is to provide a level similar to how lasers are used in construction to create an absolutely flat level. The “Level 1” dirt roads that I describe are only compacted. Sintering (using microwaves) could be used to create a Level 2 dirt road but that would entail a tremendous amount of power. Compacted dirt roads would allow full-scale rovers to drive at about 25 mph. This would allow the transport of water from a pole to an equatorial base in only about 27 hours using Tesla-like vehicles. In other words, by creating unsintered, dirt roads, fairly early on in lunar development the resources available at any point on the Moon could be made available to any other point on the Moon with a couple of days or so using automated, self-driving vehicles. This would accelerate lunar development nicely.

  35. Mapping underground structure is a well known technology that has been in operation on Earth for several decades.
    Set off a small charge and measure the reflected shockwaves from a number of ground microphones, feed the result into a computer and you get a 3D map of any cavities, density changes, big rocks, oil deposits etc under that location.

    Well within the capabilities of any project that would be doing something like this.

  36. Said lasers (and all the other equipment) would probably be solar powered anyway.

    So we are really discussing what method of concentrating sunlight would be most cost effective. And I can’t see a big, lightweight, mirror not winning that argument.

  37. An orbital mirror would have to be very big indeed, given the need to accomplish things while it was in sight of the point meant to be melted. But in lunar gravity even a 10 or 20 meter parabolic mirror could be quite light weight and moved about the surface.

  38. They need to level the surface, and solidify the top enough to prevent the rocket exhaust from excavating surface material and putting it into dangerous trajectories. I wonder if it would be worth magnetically separating meteoric iron, and making the sintered surface out of that?

  39. A big enough space mirror can focus the sunlight in an appropriately sized area and melt it. Yes, even in the lunar vacuum.

    If the location is chosen carefully (e.g. a plain), they could end up with very flat solid rock puddles. That is insta-landing pads!

  40. Perhaps sintering layer by layer with the incorporation of re-enforcing and including expansion joints might work, but that’s a far more involved method than that depicted.

  41. Sounds good to me!
    Both laser and sun would melt the rego. Seems I remember claims that micros would melt the grains together where they contact only, perhaps weaker but deeper. Perhaps good enuf for dust free product, or light loads.

  42. The laser sintering looks unrealistic to me, regolith is a good insulator, a sintered surface would need to be several inches thick, and as such a surface would be brittle would be prone to cracking. The day/night temperature extremes would make cracking and buckling insurmountable.

  43. Structural integrity of the ground will be a big problem for vertical big vehicles like BFR. Mars is probably worse than Luna due to much higher gravity. Not sure how they intend to solve that anytime soon. If the ground caves in (unevenly) under the landing legs, it will fall and fail.

    At least, that failure mode is not inconsistent with mr Musks intention to retire on Mars.

Comments are closed.