Compasses don't work on Mars, so how do you navigate? | Hacker News Reader

Compasses don't work on Mars, so how do you navigate?

kottke.org

83 pointsingve84 comments

I am surprised that no one has yet mentioned the Gyrocompass: https://en.wikipedia.org/wiki/Gyrocompass

The post mentions manually-set directional gyroscopes, but gyrocompasses are a step ahead of that; on any sufficiently-quickly-rotating planet, a gyrocompass will point you towards the geographic poles by noting the axis of precession of a gyroscope with arbitrary orientation. That's even better than a magnetic compass on Earth, since the magnetic poles do not line up exactly with the geographic poles, and the magnetic is non-uniform anyway (which is why navigational charts include notations of the magnetic deviation in different areas).

How maintenance intensive are they, though? Moving parts in or near vacuum tend to be a bit iffy (e.g. lubrication), especially for multi-year (or multi-decade) robotic missions.

There are non-moving part designs that utilise fibre-optics or laser rings. I've only seen either the physical or laser ring gyros. Not sure about which ones they'd use in space though.

Grandparent is talking about gyrocompasses, not ordinary gryoscopes though. I haven't heard of a laser-ring gyrocompass. Not sure it if that is even possible.

The laser gyroscopes do exist. They use lasers, plus the Sagnac effect, to measure rotations around on axis by interferometry.

https://en.wikipedia.org/wiki/Sagnac_effect

The gyrocompass orientation should, in principle, be obtainable by properly integrating gyroscope readings (and accelerometer? not sure if required).

You still need a way of determining the initial position/pose

I don't think so. The gyrocompass relies on gyroscopic precession. The precession itself can be obtained by integration, which gives orientation. Of course, the precision of this process may well be very low or unusable.

you don't need an initial heading measurement but you still need to know where you are

It would suffice to know the radius of the planet and it's rate of rotation.

Using only a compass and the knowledge that you were somewhere on the planet earth, could you determine the shortest route from your locations to Chicago?

Ah I thought you were referring to the gyrocompass functionality. Yes, for general navigation you need an initial position to determine your position from inertial measurement alone (as I said, you can determine orientation and latitude, but not longitude).

I would assume that once you establish direction you can just put some 3 beacons and have a local navigation system based on that.

Maybe you would know, but are MEMS gyroscopes of accurate quality to output true north data?

Specifically referring to Invensense 9150 9dof (3 accel, 3 magnetometer, 3 gyro)

Not for very long at decent accuracy. My of the hip guess is minutes. The zero rate output is +/-20 deg/s for the part you listed with RMS noise of 0.02 deg/s.

Hmm.

A group of us at our local hackerspace is building an autonomous robot. We already have that 9DoF sensor as well as a GPS and barometer.

I'm able to accurately calculate where North(magnetic) is, as well as down, and roughly what altitude I am, and synchronize them to sporadic GPS locks. Depending how useful, this could be good at providing North(true). I'll see if any sort of calibration would be useful (considering I know my lat/lon and can calculate the expected rotational vector).

I'm investigating using ORB-SLAM with ROS to also provide accurate locations of the localization.

The whole idea is that I can use this as yet another piece in a probability-location detection as well as map other things quickly and accurately.

Yeah using a compass and a magnetic field model using GPS position should let you estimate the bias from the gyro.

Celestial navigation plus accurate clocks. I'd think this would be especially good on a planet without clouds (as long as your planet rotates fast enough, so I guess Mercury is out). Anyone know if Mars has any polar stars?

https://en.wikipedia.org/wiki/Celestial_navigation

The planet doesn't have to rotate to use celestial navigation. As long as you can see the stars you are fine.

So the dark side of mercury would work fine. On the hot side as long as you are not exactly on the equator the sun acts as a fixed direction and you can use that.

> Anyone know if Mars has any polar stars?

The orbital tilt of Mars is 1.85 degrees (vs 0 for earth) and the Axial tilt is 25.19 vs 23.44 for Earth. So Polaris would be pretty close to being a polar star for Mars as well.

That's still like a couple of degrees (4 full moons) away so not sure how accurate it would be.

Still the thought of seeing literally the same stars on an alien world is blowing my mind. Just imagine the same sky, but different planet... wow

I wonder if "giant head" enhanced depth perception works across planets?

https://xkcd.com/941/

Fun link. I was planning on similar setup using a couple of remote linked Celestron C8s separated by 5 miles to observe thunderhead cloud formation in SW Florida. Average natural IPD is around 64mm so it should be interesting to see what the occipital lobe does when this is expanded to over 8km. (IPD = InterPupillary Distance; eyeball separation)

IPD?

Interpupillary distance, the distance between the centre of the pupil in one eye to the other.

Just a nitpick, but how could one see the tops of clouds when the two cameras are only negligibly above ground level?

In the flat FL interior during hot summer days with no wind it's common to observe distinct long wide columns of clouds form anvil tops over the space of a few hours in the early afternoon. Sometimes, if you're lucky, a cauldron shape will form which puts on an impressive light show at night. Rather than risking a light Cessna it might be interesting to view formation in exaggerated 3D from the ground.

Stellar parallax is virtually unnoticeable even when the Earth is on opposite sides of the Sun (~300M km apart), so I don't think the stars would look much differet even if you doubled or tripled that distance.

Objects inside of our own solar system, on the other hand, would be really interesting to look at through such a setup.

> That's still like a couple of degrees (4 full moons) away so not sure how accurate it would be.

It depends on how the two tilts interact - do they cancel or add? I don't know.

On the hot side as long as you are not exactly on the equator the sun acts as a fixed direction and you can use that.

ISTM this would work on the equator too?

it would work anywhere you could figure out the direction of the light. An example might be a tree on top of a hill, compare that to that tree's shadow on some high cliffs. That's not going to be particularly accurate, but it would get you going in the right general direction.

If you can figure out where the sun is, and you know what time it is, you're in good shape.

On the equator the sun could be directly overhead - so you have no idea what any directions are, the sun is equally far from all horizons. (Remember the sun does not move on mercury.)

If you are on the equator, but not directly under the sun then you can figure things out.

I guess I should have said equator on the central meridian.

Edit: I was under the mistaken assumption that Mercury is tidally locked.

So it would sort of work on the equator as well, but you might have to wait a bit for the sun to move enough to tell where you are.

The sun does move on Mercury. Mercury is not tidally locked- it has a 3:2 spin-orbit resonance. I.e., it rotates 3 times for every two times it goes around the sun. This means that the two sides of the planet alternate facing the sun at perihelion.

Interesting tidbit about Mercury's rotation: although it's clearly visible to the naked eye and thus has been known since before the dawn of recorded history, and has been observed through telescopes since telescopes were invented, the fact that it is not tidally locked has only been known since 1965.

Mercury is pretty featureless, right? It's not like Jupiter where you could see the rotation just by keeping track of the Great Red Spot.

It has lots of craters and such. I don't know what it looks like through an older telescope. According to Wikipedia, the problem was that because the rotation is in a resonance with its orbit, and because it's so close to the sun and thus can only be observed at certain times, it was always observed when in just one orientation.

> So Polaris would be pretty close to being a polar star for Mars as well.

Poris is only useful if you're in the northern hemisphere. In the southern hemisphere on earth we use the southern cross. It's a bit trickier if you're using a quadrant because you have to align it with a blank area of sky. Though the 17th century explorers managed it.

The pole star of a planet is in the direction of its axis of rotation. The orbits of Earth and Mars around the sun are in similar planes, and their axial tilts are similar, but they point in different directions. According to this

http://abaaonline.blogspot.com/2013/07/pole-star-on-differen...

Mars north pole points near Deneb. The reference also gives pole stars for other planets.

An easier method than creating a star map is to use the brightest star in the sky (the sun) and an accelerometer. This is the method Opportunity uses to calibrate the gyroscopes/wheel encoders.

[source pg 4] https://www-robotics.jpl.nasa.gov/publications/Reg_Willson/M...

"The attitude of the rover is based on measurements from two vector instruments: (1) accelerometers that determine the vector towards the center of gravity of Mars and (2) Pancam solar images that determine the vector to the Sun."

Star maps can be downloaded free.

Because you generally only need the brightest subset of stars in any area, you don't even need a comprehensive one.

A star map for Mars would functionally be pretty much equivalent to an earth one, I guess, given the distances involved.

The stars would be the same, of course. Out of curiosity, I launched Stellarium, moved to Mars, set myself at 90° latitude north, and watched directly overhead. Mars north pole seems to be aligned halfway between Cygnus and Perseus, so there is no "North Star" equivalent on Mars.

Doesn't Mars have clouds, though?

Dust clouds, maybe. Water clouds, no, not at all.

That's actually not correct: http://mars.jpl.nasa.gov/MPF/science/clouds.html

They're pretty much all cirrus-equivalents, granted, but they're still water (more precisely, water ice) clouds. That is, unless NASA is misinterpreting them, but seeing as they're, like, the end-all-be-all experts on Martian climate, I reckon that's unlikely.

Even barring that, however, dust clouds would also cause issues for celestial navigation.

Venus too - and this is also why neither have water - no strong magnetic field to hold the heliopause at bay, so the solar wind blows water out of the atmosphere.

Mars doesn't have hydrogen because at temperatures where water is liquid, the speed of hydrogen atoms due to temperature is greater than the escape velocity from Mars.

Wouldn't they be held by surface tension?

Surface tension is magnetic attraction between molecules. There's not enough negative magnetic attraction to hold them; they're repelled by all the other molecule nuclei.

I thought it was just pressure - the more dense molecules push the Hydrogen out of the way as gravity pulls at them.

Pressure is the collective action of the molecules / atoms flying around and smashing into things.

At "normal" temperature on Mars, hydrogen atoms have a thermal velocity [1]. That velocity is greater than the escape velocity from Mars.

A lone hydrogen atom will typically bounce around a lot in the lower atmosphere. It eventually works it's way (via random scattering) to the upper layers of the atmosphere. Once the atom reaches the upper atmosphere... it's gone. It flies away, never to return.

Keep that up for a billion years, and Mars loses most of the hydrogen it started off with.

Oxygen is heaver, so it's thermal velocity is smaller than the escape velocity.

[1] https://en.wikipedia.org/wiki/Thermal_velocity

So the Mars atmosphere is a distillation column for gasses!

Every atmosphere is a distillation column for gasses. Oxygen and Nitrogen are similar enough that they don't separate (also Earth's atmosphere moves enough to mix), but you definitely find that e.g. Radon will sink to the floor, and lighter gasses like Helium will float up to high altitude.

Does having "no strong magnetic field" mean no Ionosphere and thus no Shortwave Radio?

Interesting thought. Never considered that you have planet specific propagation.

Well, depending on frequency you would still have some traversable effects regarding radio. You would have ground wave propagation at the minimum. unfortunately, most radio would be line of sight because of no atmosphere. Limited atmospherics would provide very restricted use on a multitude of frequencies mainly lower than 6 meter. If you wanted to communicate with the other side of the body, you would have to rely on a third entity like we do on here with the moon to bounce and receive. Aside from that option you would have little/no multi bounce

Very interesting. Thanks for the detailed reply.

KRET develops stellar navigation system for Russian strategic bombers

http://www.janes.com/article/52661/kret-develops-stellar-nav...

I am seeing a cartoon forming in my head with a Russian strategic bomber pilot, crashed in a kid's bedroom and looking up at a roof covered in those glow in the dark stars and a big red flashing light warning about course deviation.

Already done? More than 40 years ago, but probably still workable:

https://en.wikipedia.org/wiki/Lunar_Roving_Vehicle#Control_a...

Did you read the article? That's exactly what the author was pointing out. Also this article and wiki have ruined The Martian for me.

Gyroscopes accumulate error with time, they make sense for an application with a short range but they're useless over long periods of time without a reference standard to crosscheck and recalibrate. That's one part of The Martian the author got entirely correct despite the criticism. The mechanical odometer in your car is only accurate to +/-3% of reading and the best mechanical calibration standards are only 0.25% of reading, that puts the trip described in The Martian off by ~90km.

Thanks, I was being mildly sarcastic in my post but having the math laid out is great.

NASA has the brains to design a better odometer but navigating by stars and GPS makes more sense.

Also, remember the +/-3% next time you're comparing a used cars. Worrying about 5-6,000 miles on a car with 100,000 is pointless since the odometer itself lies by up to 3,000 miles. Then you have further error introduced by new sets of tires which wear down 6-15mm depending on their tread depth and can vary several mm in diameter despite having the same nominal size as the OEM tires.

stellar inertial navigation - It means that as well as an inertial system you track the stars... which are actually visible even during daylight through an automated telecope system... although not through cloud I guess.

Pricey but so is getting to Mars

http://nmp.nasa.gov/st6/TECHNOLOGY/compass_tech.html

GPS

We haven't put GPS satellites around Mars quite yet. I'm guessing that won't happen until it gets terraformed.

On the other hand if we are coming in from space in the first place, it might not be too terribly hard to drop a few satellites off on your way down?

Further if you're not looking for realtime direction finding and you're willing to wait a few minutes for a fix, you can get away with far fewer satellites.

Putting 6 small satellites in orbit and getting position fixes every hour would still be way better than nothing.

How would the satellites themselves maintain their proper positions relative to Mars, without ground stations to orient against?

We already stipulated that we're landing something on the surface. That could include a stationary transmitter, which satellites could use to periodically update their location and velocity.

Yeah, agreed, my comment was meant to politely point out the necessity to the people thinking we could get away with "dropping a few satellites" ;)

All of these things can be done remotely without any wetware on Mars.

"How?"

Like we do everything on and around Mars now.

Neutron stars -- http://www.nasa.gov/topics/universe/features/zombie-stars-si...

It could be set up years ahead of any significant ground presence, at a fairly small cost.

Existing GPS satellites orient themselves by ground stations, so we'd need at least some ground presence. Said presence doesn't have to be manned, though.

Or they could orient by celestial navigation. Or have their position measured remotely and updated from Earth.

That being said, how many satellites are orbiting Mars right now that could be used to help determine position?

https://en.wikipedia.org/wiki/Category:Artificial_satellites...

Looks like 5 of those are active rather than proposed.

I think terraformed is long after we'd want GPS. Once we have any sort of Martian colony (which is usually a prerequisite for terraforming, not the other way around) and the ability to build rockets and fuel from Martian resources, it should be straightforward for Martians to set up a GPS network of their own (easier than Earth, probably, due to the lower gravity and thinner atmosphere).

The effort of getting humans to Mars will be sufficiently large that adding a few GPS satellites won't be a big deal.

Maybe one of the existing satellites could emit a radio beacon or something?

Reply on news.ycombinator.com