Hacker News Reader

Comment by durkie | Hacker News Reader

Mostly off-topic here, but it seems like the best place to get an answer: several years ago in my lab at Georgia Tech, a technician I worked with said he had a friend who invented a device for doing celestial navigation, but that it 1) worked in the daytime somehow and 2) used digital technology so that it was more precise than a sextant / traditional celestial navigation.

The technician was prone to a bit of exaggeration, so I'm not 100% sure that this is real, but I'd love to find out more if so.

View on HN

It's real. There are some neat patents from the mid-40's through the mid-60's from the usual crowd of aviation companies.

This is probably the best one: https://patents.google.com/patent/US3165632?oq=%22star+track...

The core technologies are IR filter (to increase contrast) + an optical modulator + a synchronous detector (lock-in amplifier).

There are a bunch of variations (mostly, I think, to avoid patent infringement). The basic idea is to have a sensor offset from the optical axis of a telescope (and revolving around the OA--or, more commonly, to rotate the OA about the sensor). The telescope is focused at infinity.

Then modulate the entire field in various clever ways (to avoid the gradient of the sky brightness from affecting the measurement), and detect the total field brightness with a PMT (nowdays you'd use a CMOS sensor or photodiode).

They typically include a servo to point the telescope OA at the star (conveniently, this also tells you how the star is oriented relative to the body axis of whatever the tracker is mounted on).

This is the best reference I've found so far on the SR-71 system (the NAS-14V2): https://airandspace.si.edu/webimages/collections/full/NAS-14...

That's some kind of flight/operational manual for the astroinertial navigation system (ANS), pretty detailed. Quite amazing that it's been declassified - you can see the link is from the Smithsonian Air and Space Museum. It also discusses details of how missions are loaded and planned with respect to the ANS. It also says that it has planet avoidance information, which gives you some idea of how sensitive it is (obviously the Sun and Moon are avoided).

I've not found any indication of what wavelength was used, but I would imagine the detector is a single infrared photodiode. By blocking visible atmospheric light (scattered), you can see the stars behind. Stars are point sources for all practical purposes, so you can have as large a magnification as you like without resolving them. What that means is if you have an large-aperture telescope, you'll still collect all the light from a star even if your field of view is tiny. Hence why this uses a scanning telescope to sweep around the sky.

Given the technology of the day, it's unlikely they had a multi-pixel detector (maybe a very low res array). So one approach you could use for this is two photodiodes looking slightly apart - by subtracting the signal of one from the other you might be able to correct for local atmospheric light (if the field of view is narrow enough that you don't expect two stars in the same field). Essentially a real-time dark frame. I have no idea how it actually works in practice though. I'm sure nowadays this can be done with a single wide angle camera.

It also seems like it has some strong initial conditions based on e.g. where the aircraft took off as well as references from the onboard IMU, aircraft attitude, etc. And if I read it right, the pilots can also indicate when they've reached waypoints to calibrate the system.

Very cool! Can you expand on the idea that it's just a single IR photodiode? This would get rid of visible atmospheric light, but wouldn't there still be residual IR light from the sun (whatever isn't absorbed by the atmosphere) interfering with collecting the light from stars?

Typically the FOV of the telescope is too narrow to see the sun directly, so you only have to worry about scattered light from the atmosphere (hence the IR-pass filter).

Referencing the link from my other comment, the nutator and optical modulator referenced in the patent combine to remove the residual sky background and apply FM modulation to the star aligned with the telescope OA.

By measuring the phase and depth of the FM modulation, you get the angle and distance of offset between the star and OA.

From the sibling comment seems like I had the right idea. The main restriction I was thinking of is what technology was used at the time. The SR 71 used film for imaging, and the CCD array was developed later on and probably wouldn't have enough sensitivity even if there was an early prototype. You'd also need an InGaS(?) detector for SWIR and I'm not sure when they first came out.

For this application you want as sensitive as you can get, since you want to detect minute differences in background intensity. That points to a big ole photodiode.

Hence why I suggested that they use an off axis sensor to do online dark subtraction. Locally if there was no star, the diodes should give you the same intensity. If a star enters the field of view of one, but not the other, then you have a detection. It's more complicated than that apparently, but the idea is sound.

From the patent it actually looks like they use some kind of optical chopper and a single diode. A chopper is a disc with slits cut out, in the fold of view. If there's a star that doesn't fill the field of view, the chopper will give you a periodic signal as it rotates. If the sky is empty then you'll get roughly the same signal all the way round. (from a quick skim)

They used astro-navigation as far back as the SR71, the "guidance group" system nicknamed R2D2. Cool video of a pilot talking about it here: https://youtu.be/CeBu6mRDaro?t=3180

Not just exotic aircraft! Celestial navigation was used by all kinds of long range aircraft, military and civilian, until the 1960s or so. Early jetliners (including the first models of the 747, for example) had sextant ports in the top of the fuselage; older aircraft might have "astrodomes". There's a continuous line of use from age-of-sail seafarers to jet-age aircraft. I think the military trained with sextants for air navigation until the 1990s, as a backup method for inertial and satellite.

The main difference being that star sights were taken manually, with an instrument and an eyeball, instead of automatically like the SR71.

Of course satellites did and still do use star trackers. The Apollo astronauts took star sights too, presumably for orientation rather than position.

He might have been exaggerating the performance of his friend's "invention", but you can track the sun during the day, stars at night, and possibly other celestial bodies during the day with sensitive enough instruments. If you can calculate the angles between local gravity, the height of the sun, and you know what time it is, there is a lot of location information there, especially if you track changes over time and not just instantaneous.

I think all you need is two celestial bodies and local gravity to get a fix on your location.

Don't forget that the surface of the sun is not perfectly uniform: if you have a recent picture of the surface of the sun and sensitive enough instruments, you can effectively treat it as a "cluster of known celestial objects" (e.g. sunspots, solar flares) half a degree apart from each other.

Also the moon is sometimes visible during the day, and since the topography of the moon stays constant over time that trick will work with the moon (with craters etc) even without a recent picture.

One of the most common celestial navigation methods is through a "noon sight" measuring the angle between the sun at its highest point in the day and the horizon. So that already works in the daytime.

One of the main reasons to practice celestial navigation is to have a backup in case electronic systems fail (all that salt water and all..) So I'm not so sure about using a "digital system"..

There's a recent presentation on one from NASA: https://core.ac.uk/download/pdf/286393566.pdf

This is real. It's used a lot in defense applications.

Reply on news.ycombinator.com