We are building a 3D game world that portrays an advanced lunar town that's as technically accurate as possible, for the purpose of people roleplaying living there and learning about space development. We have a complex rover right now closely based on one that was designed at JPL for the Constellation program.

enter image description here

It's still rather rough, it doesn't have much detail and the joints need redoing. It's very large, in this posture the distance between legs is around 15 m. We'll be making other more conventional vehicles but this one is good for considering the issues, I think.

The town is large and vehicles need to get around quickly on the surface. Never mind if that's how a real town would work, for the game, it's necessary. The vehicles are going to need a top speed of 120 km/h or so. Control at that speed is the issue.

I said the rover can use its legs to shift its weight to help it go around curves and brace when it brakes. That helps, but it isn't enough, and I don't know that it makes sense to do that with other vehicles. Roads can use some kind of porous sintered surface to maximize grip. Also helps, still not enough.

I've toyed with the idea of drifting - having the vehicle software do planned slides where the wheels accelerate into the direction being turned towards while working with the fact that the vehicle is going to continue in the previous direction to a large degree until the friction of the wheels succeeds in moving it in the new direction. I think it's better explained by watching a few seconds of Tokyo Drift.


That might be enough but you'd need wide roads, and to anticipate turns always, and you don't want anything unexpected. Although one nice thing is that all the wheels on the rover can turn in any direction they like at any time, and spin independently in either direction at whatever speed served navigation best.

Banking the roads a lot will help. But might that be a problem for vehicles moving slower, as it might need to be extreme? Maybe it would be necessary to also have lanes that aren't banked, for slower vehicles.

The last concept I have is to put claws on all wheels, that dig in when necessary. So road material would also need to be good for being dug into.

Can anyone help me refine these concepts into an overall system?

  • 1
    $\begingroup$ Comments are not for extended discussion; this conversation has been moved to chat. $\endgroup$
    – L.Dutch
    Commented May 6, 2021 at 19:16
  • 1
    $\begingroup$ Purely FYI you may not be aware these exist commonly on Earth! theoldrobots.com/Walking-Robot2.html $\endgroup$
    – Fattie
    Commented May 7, 2021 at 18:03
  • $\begingroup$ @Fattie no, I wasn't aware of that. Thanks! That's an interesting point of comparison. $\endgroup$
    – kim holder
    Commented May 7, 2021 at 18:13
  • $\begingroup$ Lack of friction? or lack of gravity? Ever wonder why your car has a suspension? The real purpose is not passenger comfort. The real, extremely important purpose for a road vechicle's suspension is to keep the wheels pressed against the road surface at all times. That's harder when the vehicle goes faster (at high speeds, it doesn't take as much of a bump to send the vehicle flying), and it's going to be way harder in 1/6th of Earth's surface gravity. Those long legs will help, IF they are well equipped with springs and shock absorbers. $\endgroup$ Commented May 8, 2021 at 0:08

12 Answers 12



Just run rails next to the roads, or between movement lanes.

Your vehicle clings to the rail, allowing it to zip along at much higher speed than mere surface traction allows.
This could even allow them to crest a "hill" and not fly off into the distance, which would be a very real problem when gravity is low, speeds are high and the terrain is not perfectly flat.

In effect, your vehicles move like amusement park rollercoasters.

With the difference that they are self-propelled, and their agility allows them to move from one rail to an adjacent one with ease, thus allowing navigation.

They would still be able to move over "non-road" surface, but at the vastly reduced speed that surface traction on dusty regolith will allow. Certainly much, much less than along a prepared, railled, roadlane.

  • 6
    $\begingroup$ Note that Earth trams and trains do not cling to rails but just sit on top $\endgroup$ Commented May 6, 2021 at 15:19
  • $\begingroup$ I think I'm going to go with electromagnets in the wheels sticking to iron rails. But I'm inclined to make those rails flush with the rest of the road surface so the wheels don't need a groove, making their profile better for when they aren't using rails. Make sense? $\endgroup$
    – kim holder
    Commented May 6, 2021 at 15:49
  • 2
    $\begingroup$ @user253751 yep. which is why I use the rollercoaster as example. Have the vehicle clamp itself around the rail, like a rollercoaster does. Have the wheels doing the clamping be powered. Give the vehicle the ability to change tracks while in motion. $\endgroup$
    – PcMan
    Commented May 6, 2021 at 17:55
  • 2
    $\begingroup$ @user253751 that not entirely true some trams have wheels that roll on the underside of the rail for steep slopes. look at the Translohr guide rail design. $\endgroup$
    – John
    Commented May 6, 2021 at 18:03
  • 2
    $\begingroup$ @kimholder In that case, you could use only one rail and put the magnets in a separate mechanism under the vehicle, so they're not spinning, which will make the engineering and manufacturing easier. Of course, no matter where you put them, the magnets will induce eddy currents in the rail which will act as a brake, so your vehicle will need to burn more fuel to overcome that. Which may or may not be a problem, depending on how plentiful fuel/electricity/whatever is in your world. $\endgroup$ Commented May 7, 2021 at 2:56

Basic limits

Your vehicle is limited by the coefficient of friction between the tyres and the road, multiplied by the normal force. Since both the required force to accelerate and the normal force are proportional to the mass of the vehicle, this works out to be "like Earth, but slower due to the lower gravity". However, that only applies to acceleration. Your vehicle can ultimately reach higher speeds than Earth vehicles due to the lower wind resistance.

I'll assume from now on that you want high acceleration for "cool manoeuvering".


If your road surfaces can be relied upon to be either hard-to-shift or sandy (easy to dig into), you can borrow from cold-weather wheels on Earth, and add spikes to dig into the road. This increases the lateral force you can apply, by exploiting the mass and inertia of the soil which has to be moved before the spike can move. In sandier soil, your wheels might have shaped blades on them to dig deeper and come out more cleanly, at the cost of not being able to venture onto hard ground. In game terms, this will look like bladed types, flying sand, and a slight y-offset.

Don't drift

On Earth, there are complicated anti-lock braking systems to prevent drifting. This is because drifting occurs with imperfect contact between wheel and road, or when one-or-the-other is disintegrating. Drifting actually reduces the force you can apply.


Banking the curves is possible and is used on Earth to allow inertia to substitute for gravity. At the logical extreme, a rollercoaster loop uses this - and you will notice that gravity always feels like it points "down". Yes, you would need different slopes, but a continuous curve will work better than lanes - faster vehicles will naturally travel higher up. However, there's a reason that Earth highways don't have sharp turns - the inertia of the passengers will squish them "down". This problem will be exaggerated for people used to lunar gravity. For a game, that just means your vehicle will sink lower on its simulated suspension.


Anywhere you can build a road, you can build a rail. Rails require tensile as well as compressive strength, but that's already necessary if you want your wheel to have maximum grip. Taking a high-speed tram from A to B is likely to be better for all trips except long exploration, unless the trips are short enough to walk. Magnetised tyres can also stick to a metal rail, giving players an incentive to follow the tracks with vehicles.


If you're making a game, you can declare positions to change precisely as you wish - or ramp up the ground's coefficient of friction until your lunar surface could grip greased butter. If you need it to, and it doesn't cause other problems, it's an acceptable break from reality.


JBH's answer mentions sticking a rocket on top of your car. But at that point, why point it downward? This is far from efficient or realistic, but has gameplay advantages. If you want to justify it, your car can scoop up soil and toss it out behind - hardly efficient, but nor is reaching 120 km/h in a couple of seconds for a trip so short that the acceleration time matters.


You only need lateral force, and nobody said it had to come in via the wheels... if every street /intersection has a sturdy pole, you can use that to make any turn you need. Road rules and collision avoidance are your players' problem. (Though I'd advise some "please slow down in urban areas" signage.)

Sharp corners only matter if there's something to corner around, and if there's something to turn around then there's something to grab onto. (Oh, except for chasms. "Aaaggghh!")

  • $\begingroup$ Teeth might work better the other way around, rack and pinion style. Digging into the terrain will take a lot of energy and make it hard to reach/maintain 120 km/h. Although the cloud of sharp, floating regolith should deter tailgaters with the threat of scratched paint and damaged axle bearings. $\endgroup$
    – jb6330
    Commented May 6, 2021 at 13:59
  • $\begingroup$ I think I'm going to go with electromagnets in the wheels sticking to iron rails. I like the hookshot concept though. I think that could work with a roundabout, where all vehicles entering the intersection must travel in the same direction, though disconnecting to take your exit might be more complicated in high traffic. $\endgroup$
    – kim holder
    Commented May 6, 2021 at 15:43
  • $\begingroup$ And also banking, I'll put that in too, with continuous curves. $\endgroup$
    – kim holder
    Commented May 6, 2021 at 15:44
  • $\begingroup$ I think the place I'm going to use teeth is for off-road travel. A large area around the habitats has had the top layer of regolith fused and roughened because regolith is a menace, as @jb6330 alludes too. But a few sharp teeth that deploy and dig into the ground when there is a need to brake sharply could be critical every now and then. $\endgroup$
    – kim holder
    Commented May 6, 2021 at 16:12

My first idea is a bit facetious, but velcro. Now, having said that, let's create something velcroish.

  • Let's magnetize the roadway. High power magnets are interwoven into the rubber (or whatever material you're using) of the tires. The roads are built with ferrous tracks. Basically we're holding the car onto the road, allowing for greater speed. EDIT: Or, now that I think about it, put the magnets in the road where they can have a boat load of power applied to them and make your tires magnetic-steel-belted radials. Probably a bit more efficient. Cheers.

My second idea was also a bit facetious, but vacuums. ...

  • I've often wondered if race cars develop a vacuum under the car to hold it onto the road better (beats me if they do). Extending this to the moon, where there's already a vacuum, let's add thrust to the top of the vehicle to compensate for the lack of gravity holding it down. I doubt an ion thruster would be enough, but since we combust with gas here on Earth, what's to say that a little gas (and oxidizer) wouldn't be useful on the Moon? A little illogical logic suggests that you'd get the same mileage since the fuel you're not using to push the car forward would be needed to hold it down.

My third idea is a lot less facetious.

  • We already have a thriving monorail system here on Earth - why wouldn't you use a monorail system on the moon? (I know that you have game requirements that demand personal, free-to-drive-anywhere vehicles, but I'm throwing this out there. Call it a frame challenge.)
  • 5
    $\begingroup$ Re your second point - not a vacuum, but yes, the spoilers and other surfaces on Earth-bound racing cars do create downwards pressure on the car to increase traction. $\endgroup$ Commented May 6, 2021 at 5:18
  • 2
    $\begingroup$ @KerrAvon2055 There were that formula one cars that literally had a fan on the underside of their body and functioned by "sucking" themselves down to the ground ( i think, it's been some time). More info in the wikipedia article (https://en.wikipedia.org/wiki/Brabham_BT46) in the section about the "fan car" $\endgroup$
    – mishan
    Commented May 6, 2021 at 12:21
  • 2
    $\begingroup$ For the first idea, it's probably better to use electromagnets in the wheels themselves (potentially as part of the in-hub electrical motor?), as it will be cheaper to install the hardware in 4 wheels as opposed to installing it in many kilometres of roadway. It also would allow you to disengage the magnets for if you need to drive (or push) the vehicle off the road, such as after an accident. $\endgroup$
    – jb6330
    Commented May 6, 2021 at 13:47
  • 3
    $\begingroup$ we have a thriving monorail system on Earth???? $\endgroup$ Commented May 6, 2021 at 15:19
  • 1
    $\begingroup$ @AmiralPatate "Lift on a car is... problematic" - good thing the idea never took off then $\endgroup$ Commented May 6, 2021 at 16:06

If you are going cross country then you'll want something approximating a dune buggy - which unsurprisingly is pretty much what the lunar rover looks like, which the other answers have noted.

I could imagine a suspension system that could point the axle of any wheel in any direction, under software control, which could proactively retract (raise) a wheel that's about to go over a bump to smooth out your ride and reduce the risk of going "air" borne, or point it sideways to improve cornering.

If you're building roads rather than going going cross-country, then "low friction" isn't a problem; rubber tyres will still give you roughly 1-gee (lunar, of course) of static force in any direction. (If anything, traction should be slightly better because there's no air - or water - getting between the tyres and the road.)

However maybe the real point of this question is acceleration, braking and cornering, rather than friction per-sé. In that case I would just assume that you make all the corners 6× wider, the race 6× longer, etc.

I think you're going to see a more endurance racing style rather than drag-race style.

You might want to consider having broadly 2 classes of vehicles: pressurized and open.

Rules for crash-resistance would also be a bit more stringent than on earth, as you not only have to protect the occupants from impact, but also protect their life-support systems (at least long enough for the "ambulance" to arrive).

  • $\begingroup$ Bur rubber tires won't give you 1g of static force since you're on the moon and gravity is much less than that of Earth, right? $\endgroup$
    – Cullub
    Commented May 6, 2021 at 16:42
  • 2
    $\begingroup$ I meant "g" for whatever gravitational body you're driving on. So about 1.5 N/kg on the moon. ("G" is Newton's gravitational constant, which is entirely different.) $\endgroup$ Commented May 6, 2021 at 16:45
  • $\begingroup$ I'm choosing to interpret "friction" in the question as "coefficient of friction". The alternative intuitive meaning, "force due to friction", is dependent on the weight, so you'd get more of that kind of "friction" simply by increasing the mass (which would do nothing to improve drivability, of course). $\endgroup$ Commented May 6, 2021 at 16:51
  • $\begingroup$ Yeah, I did mean for cornering and such - I didn't state that and your point is well taken. I'll try to remember that as I'm a non-technical person trying to work on a technical project. I am planning to make pretty much a dune buggy and will consider these points when I do. $\endgroup$
    – kim holder
    Commented May 6, 2021 at 16:53
  • $\begingroup$ I would substantially raise the expected maximum speed, since without air resistance, the main limits are (1) driving stability; (2) modulating the frequency of alternating electric current into the drive motors; and (3) ) "peeling" to pull the rubber up off the road surface. Maybe not as high as 720 km/h, but 300 km/h certainly seem attainable on sufficiently smooth ground. The major risk in such a race wouldn't be impact with major obstacles, but rather in failing to correctly assess the undulations and thus jumping high enough that you're not facing wheels-down when you land. $\endgroup$ Commented May 6, 2021 at 17:02

Wrong kind of wheels and tyres

If you're going any kind of speed, then baby pushchair wheels are not your friend. They're too thin so they'll snap with any kind of side loading, they have next to no contact area, and they don't roll well over anything but an ideally-smooth surface.

You want motorbike tyres. They're designed to roll well in a straight line, but also with curved walls to allow the wheel to lean and apply side loads well. Instead of "pushchair" wheels with a central mount and two skinny wheels on each side, you want "bike" wheels with a mount on each side and one fat wheel in the middle. The lower joint of your rover will basically be a bike fork. The tyre has some natural "give" to let it roll over unsmooth surfaces, and the legs of the rover presumably will become active suspension.

I'm not clear how you expect this thing to get up to speed. Will you have motors driving the wheels? Or will it essentially be a roller-skating robot, like a Wheeler with more legs?

  • $\begingroup$ Yeah, that's a good point. I can make the wheels thicker and with a good curve. The actual prototype rover this is based on has two wheels at the end of each leg, which presumably is for a reason, but I think that can be seen as a detail that several generations later isn't needed. $\endgroup$
    – kim holder
    Commented May 6, 2021 at 15:20
  • $\begingroup$ Also, motors in the wheels. $\endgroup$
    – kim holder
    Commented May 6, 2021 at 17:39

Vertical walls on curves

Put a vertical wall on the outside of each curve. The rover keeps the middle outside wheel remaining vertically oriented on the ground and raises the forward and rear legs to the horizontal position so the wheels are running along the wall. Key considerations:

  • This only works for curves, not intersections, unless you want the rover to be using its legs to jump over walls when it wants to go straight ahead! Maybe use the magnetic system proposed by JBH as part of the traffic control at intersections only without needing to magnetise all sections of all roads.
  • By corollary, roads can be relatively crude surfaces so long as they are smooth - hitting even a very mild bump at 120 kph at 1/6 G will send a rover soaring out of contact with the road surface.
  • Height of the wheels running along the wall (ie the practical wall height) needs to be as high as the centre of gravity of the vehicle to allow for considerable speed.
  • Practical engineering consideration - given the specified top speed of 120 kph, build the walls and legs so they can take the outwards force they will experience at that speed.

One disadvantage of a wall with wheels rather than banking the road is that significant sideways force will be experienced by the occupants and cargo of the rovers - secure payloads accordingly.

One further general consideration is emergency braking for cars travelling at 120 kph on the moon. Braking by friction with the road will be very slow. Parachutes (as used by drag racers on Earth) are obviously no use. Retro rockets are the best bet for emergency braking, firing anchors requires very careful engineering of cables and will have variable results depending on the surface the anchor is interacting with.

  • $\begingroup$ the vehicle illustrated appeas cabale of tilting the cabin to any angle allowing cornering forces to be ecperienced in a downwards direction. $\endgroup$
    – Jasen
    Commented May 7, 2021 at 22:45


Since you are in a vacuum, and in low gravity, a maglev track needs to spend far less energy to keep the train elevated, and does not need to worry about drag. Additionally in the vacuum environment, you may be able to keep the superconductors cold without excessive effort, so long as they are not attached to anything warm, as the vacuum forms an insulator.

  • $\begingroup$ Maglev would certainly work well on the moon. Right now though, this is for the kind of thing where more flexibility is needed. Vehicles are also a more attractive option for individual gameplay. $\endgroup$
    – kim holder
    Commented May 6, 2021 at 16:35
  • $\begingroup$ Maglev doesn't have to be a train, you could have single cars on a maglev track. I'd also take a look at youtube.com/watch?v=8OBHUZetYIU for some options on how flexible you can be with linear magnets. $\endgroup$ Commented May 6, 2021 at 16:38
  • $\begingroup$ Yeah, fair point. Still, players are going to use vehicles for exploring the surface. They'll want to go wherever, and whatever speed they choose. I bet we'll use maglev at some point in this project, but not here. $\endgroup$
    – kim holder
    Commented May 6, 2021 at 16:45
  • $\begingroup$ Presence of a vacuum has little or nothing to do with temperature. The surface of the moon ranges from -173C at night (colder in some southern craters), to 127C in daylight. The lack of atmosphere precludes convective heat dissipation. Given the extreme conditions, the equipment required to remove excess heat from lunar superconductors would in fact be a bit more complicated/expensive than earth bound equipment. $\endgroup$
    – jwdonahue
    Commented May 6, 2021 at 20:58
  • 1
    $\begingroup$ Whilst presence of a vacuum has minimal impact on temperature, for something like superconductors, the presence of high quality vacuum has a huge impact on your ability to insulate them from other heat sources. And once they are cooled down, they do not produce much heat of their own. $\endgroup$ Commented May 6, 2021 at 21:05

Larger curve radii and larger braking distance

Being able to control a car at 120 km/h isn't easy in the Earth, either, so we need large highways with smooth slopes, long acceleration lanes and large curves. In the Moon that isn't different, but since friction force is about one sixth of that in Earth, for the same design speed curve radii should be about six times larger than here. Acceleration and braking lanes should be about six times longer.

Other answers have mentioned banking. Banking six times more than in Earth highways don't seem to have to cause problems, although the exact ratio depends on the combination of radius and speed.

And just a side note: beware of dust. Lunar dust might affect the friction coefficient between wheel and highway. Although I can't be sure about if it is going to increase or decrease it, that would be worth investigating before designing your highway.

  • $\begingroup$ Yes, regolith dust is a menace in general. The surface of all roads would be sintered to a depth of at least 10 cm. The specifics need further analysis. $\endgroup$
    – kim holder
    Commented May 7, 2021 at 16:56
  • $\begingroup$ @kimholder Well, the avantatge is that in a hard road you can get rid of dust by brooming. A different question would be how to build a large sized hard road on the Moon. $\endgroup$
    – Pere
    Commented May 7, 2021 at 17:16

Softer rubber, terestrial automobile tires are a fairly hard rubber that grips poorly.

Using a softer rubber with a higher coefficient of friction would cause faster wear, but with the reduced gravity there will be less load on them. Drag racers use soft tires to get more grip.

Banked corners will help too. see "indy500" and "wall of death" for extreme terestrial examples

If you're going off road sling-shotting around, or hopping against fixed obstacles may help with rapid cornering. Your machine has long legs that with the right spring-loaded hinges could do quite an effective sideways hop.

  • $\begingroup$ 'soft rubber' On the moon? I will take it to mean a radically different material with similar properties. Vacuum, weeks of continuous daylight ( with shade), weeks of continuous dark is a rather harsh environment. Strikes me as being a rather significant engineering problem. $\endgroup$ Commented May 7, 2021 at 20:56
  • $\begingroup$ @GaultDrakkor Silicone does okay - ntrs.nasa.gov/api/citations/20100029591/downloads/… The application here is a lot different, but this could be possible I think. Substances containing carbon could be expensive to use in a way where they get worn away and the carbon can't be recovered, though. $\endgroup$
    – kim holder
    Commented May 8, 2021 at 20:06

Tubes of lunarcrete

At speeds on 120 km/h and higher, the equivalent of motorways on Earth, a manufactured road system on Luna would ideally consist of single-direction tubes with bendy curves, with branches but no crossroads (multiple levels instead, like motorway intersections on Earth). A system of arteries, more or less.

The cars could resemble formula racecars, with wide wheelbases and low centers of gravity. If the curvature of the tube is smooth enough overtaking faster vehicles can pass almost overhead, riding upside down for a second. Some markings on the inside of the tube are probably needed, otherwise it would be very hard to tell up from down. The inertial accelerations will be dominant, so passengers and cargo will need to be secured in all directions (multiple seatbelts, etc.). If driving upside down inside the tubes is too scary, car designs with triangular wheelbases could be trialed.

The tubes could be pressurized or not. Fully covered ceiling probably also evens the temperature conditions inside, and protects from smaller meteors, even without pressurization.


Travel by jumping

This robot/rover seems to be able to jump, and to absorb and dampen jumps, thanks to the large clearance of its legs.

Moving on the moon by jumps can have the advantage of freedom of exploration.

There could be two type of jumps:

Small electrically powered hops (1m to 20m distance, max height say 20m) could be powered by the legs only.

Bigger planned jumps assisted by a thruster lit at say 20 m of altitude, both for the departure and the arrival. (in order not to raise too much dust) The initial jump and the reception are assured by the legs, the same way as for non assisted jumps.

The thruster does not allow the vehicle to go orbital and provides gentle acceleration and deceleration, whatever the length of the jump. (say max 2g) Anyway large suborbital distances can be possible.

This solution cannot be all electric since it requires a main thruster and maybe attitude control thrusters. Yet when out of fuel, small electrically powered hops are still a way to get back to the base.


Flat roads, with lateral guard-rail (emphasis on "rail").

The vehicle moves at whatever speed it needs to: when necessary it extends two lateral legs and connects with the guard-rail (it could be a vertical border with a deep horizontal groove). Then, centripetal force is supplied by the mechanical reaction from the groove.

Two counterlateral wheels could be shaped in such a way to optimize connection with the lateral grooves at the sides of the road, using whichever of the two is more convenient. Or if necessary, the road could be exactly as wide as the vehicle with both lateral balancers extended, so that both grooves could be used - in that case, overtaking wouldn't be possible of course.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .