"The second approach was to use a digital computer to determine the solution. This solution was rejected because in 1963, a digital computer was expensive, slow, and less reliable."
This inflection point between analog and digital computer is a fascinating one. At one point in time a analog computer made sense and some later point in time you would be foolish to specify anything other than a digital computer. But that time between when it could go either way is interesting. There is a good autobiography by the person responsible for introducing the first digital computer to the navy that provides an interesting view into this era. https://ethw.org/First-Hand:No_Damned_Computer_is_Going_to_T...
Now I am vaguely searching for a guide on gear train schematic diagrams, I am sure they had them, you don't reason out something this complicated without one. I know hydraulics has it's own flavor of schematic diagram, which are fascinating if all you have seen are electronic circuits. https://www.hidraoil.com/technical-resources/hydraulic-symbo...
That is why I have doubts about their scalability. There are fundamental physics reasons we were able to scale digital computers and were unable to do the same to analog ones.
That's certainly true. But - the fastest possible machine to compute the sound of a guitar string reverberating in a concert hall is a guitar in a concert hall. The world itself does a tremendous amount of compute - the question is whether it is useful compute. I don't think this has been explored nearly enough.
Nonlin ar partial differential equations through a large continuum are expensive. Even if you can scale this particular example, it doesn't refute the point, the universe does a tremendous amount of compute that we don't know how to exploit.
I think it is totally possible for the computing to take non-zero time, but we observe it in zero-time as our consciousness only steps forward only with each iteration of computing the world state. So we observe zero time reality computations.
> No Damned Computer is Going to Tell Me What to DO
That is the best title for a story about replacing analog and mechanical instruments with digital computers. A similar process is happening now with natural intelligence, replacing or augmenting the human intellect.
I said it because, the behavior of LLMs only appear to mimic intelligence only very superficially. Even without going "behind the curtain", and by only examining the behavior of LLMs slightly deeply, the illusion of this intelligent behavior break down.
Intelligent behavior need to be consistent. You need only a single instance of non-intelligent behavior to prove that an entity is not intelligent.
This is from the era of devices where the I/O was entirely electrical but the computation was mechanical. Most of this stuff came from naval gunnery. The naval "fire control tables" started out as mechanical computers where a rather large number of people were inputting different sensor readings via cranks and dials.[1] Gradually, more of the inputs came in directly from the sensors, and more of the outputs went directly to the gun turrets. The final form of this technology was units the size of a footlocker full of gears, cams, and resolvers, with all-electric inputs and outputs.
Such things used to show up in surplus stores.
I've seen the restored guidance computer for the Nike missile, at the site in Marin County.[2] That's similar, although ground-based. Analog data came in from radars, was processed with mechanical computation, and control signals went out to the missile.
See [1] for the basic mechanical components. It's a better scan of the same film the Periscope Film archive sells, which is the first one linked above. No sprocket clatter.
I forget where I saw it (probably YouTube someplace) but the fire control systems (including radar) on the Iowa-class battleships apparently way outperformed their Japanese counterparts. The Japanese has a couple (?) ships with bigger guns/longer range but they couldn't actually take advantage of that longer range to hit big US ships.
Haven't been there in years but the Nike facility in Marin is well worth a visit if you're there when it's open. The control stations were originally on a higher ridge but they have one of the (basically) containers next to the missile sites now. The idea at the time is that they would explode ordinance (originally conventional, later nuclear) above incoming bombers causing a pressure wave that would make them crash.
Was also a Nike base on Angel Island but there's nothing left there but some old concrete pads.
We actually had one of the Nike bases defending Philadelphia literally next to where I grew up. Don't remember personally--was very young--but there were apparently troop manoeuvres on our property from time to time.
If you're looking for more, the book "Between Human and Machine: Feedback, Control, and Computing before Cybernetics" is a detailed history of the development of electromechanical fire control computers and feedback systems.
Everytime I read articles like that, I envy the engineers that worked in development of such tools. First microprocessors in jet fighters, electromechanical celestial navigation...
I think the opposite. Hardware is hard, as they say. Building such complex electromechanical designs to military specs without modern CAD tools must have been the equivalent of writing code in binary, without high level languages or even assembler.
I’m sorry. I’m in a bad mood and that was unecessary. That being said, given the current hyper militarized climate in Silicon Valley, I find this detachment of the science / engineering from its use cases to be more than a little distasteful.
You are to be commended for an apology, it shows class and decency.
As for the militarization of Silicon Valley, it's been said we have god-like tech, but not the emotional discipline for such responsibilities. Aside from the fact that we humans suck, we repeat our worst mistakes without, it seems, a second thought. Then, when we're called out, we let our ego warp to any excuse that will suffice. The Kissinger example mentioned above almost made me ill.
another real fact: "Between 1964 and 1973, the United States conducted a covert "Secret War" in Laos, dropping over two million tons of ordnance during 580,000+ bombing missions, "
The revelation of secret bombing campaigns was one of the main reveals of The Pentagon Papers by Daniel Ellsberg. This arguably turned American public opinion against the war decisively as it revealed the USA had no cohesive strategy for winning and was repeatedly lying to the American public about the multiple fronts of the war in Southeast Asia for more than twenty years.
Eh, it's easy to get caught by the romanticism of working on things like this, but I assure you besides like 4 people in charge of the big picture, everybody else is dealing with things which are exactly as mundane as things these days. Like putting it through 1000 heat cycles of -40 to 200 degrees and then vibrating it at 2gs for 200 hours and then measuring the tolerances of each part... or being in charge of three lines in a standards document for 2 years negotiating the details with the DoD.
I couldn't find the specification for the Angle Computer, but I've found specifications for other devices and you're exactly right: pages and pages of vibration requirements, fungus resistance, testing procedures, and then maybe if I'm lucky one page with useful information like the pinout. This is very annoying if I'm paying by the page. :-)
It's a shame the only way to work on problems like these (and make a decent living) is to make tools of war.
The end game of much of silicon valley seems to be government (read: military) contracts. Probably because its the main branch of government that's thoroughly funded
I'm shooting from the depths of my memory, but I recall reading that one of the earliest government needs for computers was for the decennial census. At some point, it was requiring more than the 10 years to process the previous censuses (sp?) results.
Read every word. i liked this detail in the footnotes:
> The Astro Compass needed to know approximately where in the sky to find the star, in order to point its sensor in the right direction. The direction didn't need to be exact because the Astro Compass performed a spiral search pattern to find the star. This search pattern covered ±4° in bearing and ±2.5° in altitude. In comparison, the Moon is 0.5° wide, so it's a fairly large target area. ↩
Laser trackers (used for metrology) can also use spiral search to find retroreflectors. Although I believe newer models generate a flash of infrared and find them via bright spots in the resulting image from a camera. I imagine modern celestial navigation systems evolved similarly (minus the infrared flash, not very useful for stars).
Honestly that footnote really stood out to me too! the spiral search detail makes the whole system feel a lot more alive than I expected like it’s actively hunting for the star rather than just pointing and hoping.
> The Atro Tracker also has declination limits of +90° and -47° and a lower altitude limit of -6°. The latitude is limited to the range between -2° and +90°; the system automatically switches hemispheres so both the North and South latitudes are usable.
Why would the system need to have a much greater range of declination (celestial sphere) than latitude (Earth spheroid)? Because the Astro Tracker and Angle Computer could flip over to the Southern hemisphere (was this automatic or was there a switch?) having that much declination range seems unnecessary. Perhaps to allow for pitch of the aircraft in flight?
BTW, being able to operate in both the Northern & Southern hemispheres was an important capability for the B-52. Previous bombers (B-36 mostly) had the range but not the reliability or in-flight refueling for global reach.
Sadly, I didn't get the chance to look at the B-52 at the Museum of Flight when I was there. If you ever meet Charles Simonyi, please thank him for his support of the museum.
If you're flying in low latitudes, nearly half the stars that you want to use are going to have negative declination, so negative declinations are important. As for the hemisphere switching, this happened automatically.
It's totally normal to be in the northern hemisphere and looking at stars below the celestial equator. For instance, Sirius is the brightest star in the night sky and is in the southern half of the celestial sphere. So if you wanted to navigate with Sirius, the system had to support negative declination. (They define negative declination as in the opposite N/S hemisphere from the aircraft.)
The B-52 is one of my favorite aircraft, and the one at the Museum of Flight is an absolute beast -- I never thought it was small, but it's still bigger than I expected.
As I understand it, the star altitude is measured relative to an artificial horizon.
How did it determine "down" in a moving airplane? Was it essentially doing the high-tech equivalent of dangling a rock on a string with some dampening (in a gyroscopic cage to avoid being affected by the airplane's rotation), or something smarter?
When I looked into whether astronavigation would be solvable cheaply or somehow trivially using modern hardware, I found this a surprisingly difficult problem even on a static platform - inclinometers that would get you down to 0.01° accuracy (which would still translate to a ~1 km positional error if I'm not mistaken, roughly what a skilled sailor is supposed to be able to do with a sextant) are expensive even today.
With a moving, shaking platform that's changing position (i.e. a perfect gyro will point perfectly in the wrong direction after a few minutes of flight) and might be flying turns (which makes "down" point in the wrong direction) that seems hard to solve.
The B-52 star tracker used a gyroscope to determine vertical. The Astro Tracker was stabilized by a bunch of motors and synchros so it matched the gyroscope. Thus, the Astro Tracker was a stable platform even as the aircraft pitched and rolled. (Footnote 4 in my article shows the vertical gyro attached to the Astro Tracker.)
> Was it essentially doing the high-tech equivalent of dangling a rock on a string with some dampening (in a gyroscopic cage to avoid being affected by the airplane's rotation), or something smarter?
Yes, that is essentially how a gyroscopic artificial horizon works.
Consider that the local horizon changes relative to an inertial frame (the stars) as you travel across the surface of a sphere, so even if you could build a perfect gyro that remained stationary in the inertial frame you would need to update the local down as you move. The solution is to slightly weight the gyro cage to bias it to the local down.
Now, consider that, in a properly-coordinated turn, the passengers (and gyro) will feel that gravity points straight to the floor :) So the time-constant of the damping is important.
I assume the constant is usually chosen short enough that the system will "forget" turns quickly, in exchange for becoming useless while turning?
Still, getting this whole thing accurate to probably one minute of arc is insane, especially with the gyro and star tracker linked only via motors and synchros. So the total error is the sum of any deviation of the gyroscope from the actual down direction, the error in measuring the gyro angle, the error in setting the star tracker to that exact angle, and then all other errors the system introduces. Then you need to take multiple separate measurements at different times and compensate for the movement, and a one-degree difference means you're over the wrong city (or in Europe, country) so the end-to-end accuracy must be much better than that.
And sailors supposedly did that with a sextant to something like 0.01° on a moving ship.
Was the star tracked manually by the navigator (as in, did they have to manually “look for” and keep track of it)? Fascinating article, but I’m not grokking how it was used in practice.
The device has a spiral search mechanism to find the star. Then it locked onto the star and automatically tracked it. So this was unlike the Apollo star tracker where the astronaut has to manually aim at the star.
I'll probably write another article on the star tracker itself. But I can give you a quick summary of the spiral search mechanism. It was electromechanical: a motor turned a resolver, a device with coils to generate sine and cosine from the shaft angle. This gives the X and Y deflections for a circle. These signals went through potentiometers that were also turned by the motor to produce constantly growing magnitudes, so you get a spiral. But you need to slow down the motor as you spiral outwards since you're covering a much larger linear region. So the motor also turns a stepping switch that progressively reduces its speed.
Once the system finds a star, a complicated feedback mechanism keeps it locked onto the star. There is a spinning slotted disk in front of the photomultiplier tube. If the star is off center, the output will peak when the slot lines up with the star. Thus there is an error signal with phase that indicates the direction to the star. This signal is demodulated to produce X and Y signals that change the aim to move towards the star.
I would absolutely love to read something about that - thanks for putting in the work and sharing it.
I have a buddy working on restoring a set of binoculars that were attached to the Target Bearing Transmitter system for a US sub from the 50s. Last I heard he was able to find someone that actually had parts of the original schematics for it so that he’s able to machine some new pieces.
Am I right in thinking it didn't matter which star it locked onto, and it didn't need to know which star it was? Would it be a problem if it locked onto another celestial body (e.g. Venus)?
No, it needed to lock onto the right star, the one that matched the coordinates. Otherwise, it would be pointing in a random direction. The navigator would check against three different stars to detect an error.
The system could also use planets or even the sun for navigation. A special filter was used with the sun to avoid burning out the photomultiplier tube.
Ah, so it could be used in the daytime. I read the whole article assuming it was only useful at night. (When else would you be flying a bomber and need high accuracy?)
Yes, haze and clouds were a problem at low altitudes, but most of the time the aircraft was above the clouds. The Aurora Borealis (northern lights) was potentially a problem; the system included an aurora filter.
Since the article doesn't mention: I've read that ICBMs used celestial navigation. Is this similar to what contemporary missiles used? Do we even know at this point?
This is crazy impressive ... the kind of thing that should inspire one to do more, much more, than whatever "mere plumbing" one happens to be doing at the moment
this is exactly what i needed to read when i am starting a mini project to turn empty Chewy and Amazon boxes into a new cat maze for my bonded pair of shelter fearsome beasts
can i do something with a v1 raspberry pi and myriad idle laptops and gadgets. both Opus 4.7 and i have had enough of each other for a Caturday
thanks and i don't know why i got downvoted. there is a corporeal world. plus one of my critters figured out how to quasi open a cabinet she thinks can access food way above. just wait until she develops opposable thumbs
> AI statement: I didn't use AI to write this article (details).
Meta, but thank you for including this and suggest even putting it at the top of your articles. I'm now off to bother to read something that someone bothered to write :)
> The Angle Computer is one piece of the Astro Compass, a system that locked onto a star and produced a highly accurate heading (i.e., compass direction), accurate to a tenth of a degree.
I think it provides ground track information not just heading? Which is far more valuable for aircraft navigation, because the main issue is unpredictable wind drift.
> The diagram below shows the guidance system of the Minuteman III missile (1970). This guidance system contains over 17,000 electronic and mechanical parts, costing $510,000 (about $4.5 million in current dollars). The heart of the guidance system is the gyro stabilized platform, which uses gyroscopes and accelerometers to measure the missile's orientation and acceleration.
Even nuttier is the one from the Peacekeeper. Float a perfect beryllium sphere in fluorocarbon. Use thrusters to keep it oriented. No gimbal lock, because no gimbals. Six million dollars per unit, in 1987. So good that a system with literally perfect accuracy wouldn't improve accuracy, because error from the system was already well below other sources of inaccuracy in the missile. https://en.wikipedia.org/wiki/Advanced_Inertial_Reference_Sp...
I don't know specifically about Minuteman though a different ICBM uses inertial guidance but also does a star sight to calibrate at some point in its trajectory--says the Internet ahem. And, also yes, the modern inertial navigation that goes into all this is pretty amazing. The not so modern iteration of all this did get us to the moon in 1969.
It's amazing, the things that can be done without what we would consider modern technology.
The 8-bit Guy recently released a video asking "What if everything still ran out vacuum tubes?" <https://www.youtube.com/watch?v=mEpnRM97ACQ>. Conclusion: A surprising amount of things we take for granted today would still be possible.
Reminds me a bit of this[0]. I have an iOS app[1] that models its operation. Sextants[2] are damn clever devices, and have been around for about three hundred years. Theodolites[3] are even older, but are used for terrestrial measurements.
Fun! I was just reading about the star tracker in "Skunk Works: A Personal Memoir of My Years at Lockheed". Really fascinating when you're thinking about how this all happened in the 50's and 60's.
In a very similar vein, Ars Technica did a very interesting story on the electromechanical targeting computers on WW2 battle ships a few years ago; the instructional videos embedded in the story are gold.
Someone recreating this and allowing access to it sort of in the style of an escape room business would be pretty cool - motion flight sim where you can learn to fly the plane or learn to operate the other parts of engineer/bombing/navigation etc. And maybe not simulating the problematic "let's bomb human targets" but rather just bullseyes in fields.
"The second approach was to use a digital computer to determine the solution. This solution was rejected because in 1963, a digital computer was expensive, slow, and less reliable."
This inflection point between analog and digital computer is a fascinating one. At one point in time a analog computer made sense and some later point in time you would be foolish to specify anything other than a digital computer. But that time between when it could go either way is interesting. There is a good autobiography by the person responsible for introducing the first digital computer to the navy that provides an interesting view into this era. https://ethw.org/First-Hand:No_Damned_Computer_is_Going_to_T...
Now I am vaguely searching for a guide on gear train schematic diagrams, I am sure they had them, you don't reason out something this complicated without one. I know hydraulics has it's own flavor of schematic diagram, which are fascinating if all you have seen are electronic circuits. https://www.hidraoil.com/technical-resources/hydraulic-symbo...
They're making a comeback now, quantum computers are analog
That is why I have doubts about their scalability. There are fundamental physics reasons we were able to scale digital computers and were unable to do the same to analog ones.
That's certainly true. But - the fastest possible machine to compute the sound of a guitar string reverberating in a concert hall is a guitar in a concert hall. The world itself does a tremendous amount of compute - the question is whether it is useful compute. I don't think this has been explored nearly enough.
Why is it the fastest possible?
Wouldn’t you be able to run a higher clock-speed simulation, and therefore compute the sound ahead-of-time, on a digital device?
Nonlin ar partial differential equations through a large continuum are expensive. Even if you can scale this particular example, it doesn't refute the point, the universe does a tremendous amount of compute that we don't know how to exploit.
This assumes that the universe is computing.
But there is some hypothesis like MUH that sees reality as a sort of "recording" that we just experience.
>Why is it the fastest possible?
Why? because the "universe" takes zero time to compute it. In fact all of reality is computed in zero time. How can you beat it?
TO continue the above comment..
I think it is totally possible for the computing to take non-zero time, but we observe it in zero-time as our consciousness only steps forward only with each iteration of computing the world state. So we observe zero time reality computations.
> No Damned Computer is Going to Tell Me What to DO
That is the best title for a story about replacing analog and mechanical instruments with digital computers. A similar process is happening now with natural intelligence, replacing or augmenting the human intellect.
An interesting resource I just found:
The analog computer museum - https://www.analogmuseum.org/english/
It has a Library section with lots of downloadable articles in German and English.
> A similar process is happening now with natural intelligence
LLMs are not AI.
Correct. Neither are computers, analog machines, or mathematics.
I'm on your side, human comrade. No damn computer is going to tell me what to do!
I said it because, the behavior of LLMs only appear to mimic intelligence only very superficially. Even without going "behind the curtain", and by only examining the behavior of LLMs slightly deeply, the illusion of this intelligent behavior break down.
Intelligent behavior need to be consistent. You need only a single instance of non-intelligent behavior to prove that an entity is not intelligent.
This is from the era of devices where the I/O was entirely electrical but the computation was mechanical. Most of this stuff came from naval gunnery. The naval "fire control tables" started out as mechanical computers where a rather large number of people were inputting different sensor readings via cranks and dials.[1] Gradually, more of the inputs came in directly from the sensors, and more of the outputs went directly to the gun turrets. The final form of this technology was units the size of a footlocker full of gears, cams, and resolvers, with all-electric inputs and outputs. Such things used to show up in surplus stores.
I've seen the restored guidance computer for the Nike missile, at the site in Marin County.[2] That's similar, although ground-based. Analog data came in from radars, was processed with mechanical computation, and control signals went out to the missile.
[1] https://en.wikipedia.org/wiki/Admiralty_Fire_Control_Table
[2] https://www.nps.gov/goga/nike-missile-site.htm
There are some old training videos that show how this worked: https://www.youtube.com/watch?v=gwf5mAlI7Ug
Also the Battleship New Jersey YouTube channel has some nice content on this: https://www.youtube.com/watch?v=szxNJydEqOs
See [1] for the basic mechanical components. It's a better scan of the same film the Periscope Film archive sells, which is the first one linked above. No sprocket clatter.
[1] https://www.youtube.com/watch?v=s1i-dnAH9Y4
I forget where I saw it (probably YouTube someplace) but the fire control systems (including radar) on the Iowa-class battleships apparently way outperformed their Japanese counterparts. The Japanese has a couple (?) ships with bigger guns/longer range but they couldn't actually take advantage of that longer range to hit big US ships.
Semi-related but people should look at the Sprint missile if they're interested in this, it goes so ridiculously fast the warhead begins to glow.
https://youtu.be/kvZGaMt7UgQ
One of my favorite internet links is an archive of manuals from this era. Especially the Torpedo Data Computer, another fire solution calculator.
Excellent illustrations!
https://maritime.org/doc/tdc/index.php
Haven't been there in years but the Nike facility in Marin is well worth a visit if you're there when it's open. The control stations were originally on a higher ridge but they have one of the (basically) containers next to the missile sites now. The idea at the time is that they would explode ordinance (originally conventional, later nuclear) above incoming bombers causing a pressure wave that would make them crash.
Was also a Nike base on Angel Island but there's nothing left there but some old concrete pads.
We actually had one of the Nike bases defending Philadelphia literally next to where I grew up. Don't remember personally--was very young--but there were apparently troop manoeuvres on our property from time to time.
There were fourteen Nike sites in the Bay Area.
We may see large numbers of local defense sites again, against drones and medium-range missiles. Israel and Iraq already have that.
If you're looking for more, the book "Between Human and Machine: Feedback, Control, and Computing before Cybernetics" is a detailed history of the development of electromechanical fire control computers and feedback systems.
Off topic, but this is where I see AI going. A tool that condenses work down from requiring a team and a room to a box. We're decades away from that
Everytime I read articles like that, I envy the engineers that worked in development of such tools. First microprocessors in jet fighters, electromechanical celestial navigation...
And here I am fighting gitlab pipelines.
I think the opposite. Hardware is hard, as they say. Building such complex electromechanical designs to military specs without modern CAD tools must have been the equivalent of writing code in binary, without high level languages or even assembler.
I'll take it any day if I wouldn't need to see docker on windows anymore.
It seems GP was appreciating the beauty and purity of engineering required in solving such problems (as opposed to, perhaps, fighting AI slop today).
Nothing is stopping us.
One life to experience the universe. Save up for a sabbatical. Find new engineering pastures.
It's always rose colored looking back. Not everybody got to work on this. Some people were storming the beaches...
And some people, specifically Vietnamese and Cambodian civilians, were on the receiving end of your fun little brain teaser.
And other people, like Henry Kissinger, drew random dots on a map to tell it where to drop the bombs. https://en.wikipedia.org/wiki/Operation_Menu
> And some people, specifically Vietnamese and Cambodian civilians, were on the receiving end of your fun little brain teaser.
To make it ABUNDANTLY CLEAR, I was referring to celestial navigation.
I guess we have to blame people who weren't alive at the time for wars we didn't participate in?
My wife is Vietnamese, btw.
I’m sorry. I’m in a bad mood and that was unecessary. That being said, given the current hyper militarized climate in Silicon Valley, I find this detachment of the science / engineering from its use cases to be more than a little distasteful.
You are to be commended for an apology, it shows class and decency.
As for the militarization of Silicon Valley, it's been said we have god-like tech, but not the emotional discipline for such responsibilities. Aside from the fact that we humans suck, we repeat our worst mistakes without, it seems, a second thought. Then, when we're called out, we let our ego warp to any excuse that will suffice. The Kissinger example mentioned above almost made me ill.
Fun fact, that's how it started, WW2 and the need for electronic warfare and countermeasures created/set the stage for silicon valley.
https://youtu.be/ZTC_RxWN_xo?si=fKlacQAM8krrW17B
another real fact: "Between 1964 and 1973, the United States conducted a covert "Secret War" in Laos, dropping over two million tons of ordnance during 580,000+ bombing missions, "
I’m about to read King Leopold’s Ghost. Great choice in username.
I must say it’s a little disappointing that things like “secret bombing campaigns” getting declassified don’t lead to much public response.
The revelation of secret bombing campaigns was one of the main reveals of The Pentagon Papers by Daniel Ellsberg. This arguably turned American public opinion against the war decisively as it revealed the USA had no cohesive strategy for winning and was repeatedly lying to the American public about the multiple fronts of the war in Southeast Asia for more than twenty years.
Eh, it's easy to get caught by the romanticism of working on things like this, but I assure you besides like 4 people in charge of the big picture, everybody else is dealing with things which are exactly as mundane as things these days. Like putting it through 1000 heat cycles of -40 to 200 degrees and then vibrating it at 2gs for 200 hours and then measuring the tolerances of each part... or being in charge of three lines in a standards document for 2 years negotiating the details with the DoD.
I couldn't find the specification for the Angle Computer, but I've found specifications for other devices and you're exactly right: pages and pages of vibration requirements, fungus resistance, testing procedures, and then maybe if I'm lucky one page with useful information like the pinout. This is very annoying if I'm paying by the page. :-)
It's a shame the only way to work on problems like these (and make a decent living) is to make tools of war.
The end game of much of silicon valley seems to be government (read: military) contracts. Probably because its the main branch of government that's thoroughly funded
I'm shooting from the depths of my memory, but I recall reading that one of the earliest government needs for computers was for the decennial census. At some point, it was requiring more than the 10 years to process the previous censuses (sp?) results.
its also a branch of government that always need research so government contracts are plentiful
Defense also got silicon valley started. So it goes full circle.
I’m with you. The complexity yet simplicity of these mechanical devices is fascinating.
> First microprocessors in jet fighters
Don't get me started on that...
But let's say I got you started. What would you want to say about them?
The short answer is that it shows what one can get people to believe through relentless self-promotion. For a longer answer on the first microprocessors: https://spectrum.ieee.org/the-surprising-story-of-the-first-...
Would you consider writing a computer history book?
Maybe at some point. The main problem is that writing a book would require me to focus on one topic, which doesn't seem likely.
Please do <3
Easy, just write multiple books simultaneously /s. Cheers.
Read every word. i liked this detail in the footnotes:
> The Astro Compass needed to know approximately where in the sky to find the star, in order to point its sensor in the right direction. The direction didn't need to be exact because the Astro Compass performed a spiral search pattern to find the star. This search pattern covered ±4° in bearing and ±2.5° in altitude. In comparison, the Moon is 0.5° wide, so it's a fairly large target area. ↩
This is also used in laser communication systems to find your peer.
Laser trackers (used for metrology) can also use spiral search to find retroreflectors. Although I believe newer models generate a flash of infrared and find them via bright spots in the resulting image from a camera. I imagine modern celestial navigation systems evolved similarly (minus the infrared flash, not very useful for stars).
Honestly that footnote really stood out to me too! the spiral search detail makes the whole system feel a lot more alive than I expected like it’s actively hunting for the star rather than just pointing and hoping.
> The Atro Tracker also has declination limits of +90° and -47° and a lower altitude limit of -6°. The latitude is limited to the range between -2° and +90°; the system automatically switches hemispheres so both the North and South latitudes are usable.
Why would the system need to have a much greater range of declination (celestial sphere) than latitude (Earth spheroid)? Because the Astro Tracker and Angle Computer could flip over to the Southern hemisphere (was this automatic or was there a switch?) having that much declination range seems unnecessary. Perhaps to allow for pitch of the aircraft in flight?
BTW, being able to operate in both the Northern & Southern hemispheres was an important capability for the B-52. Previous bombers (B-36 mostly) had the range but not the reliability or in-flight refueling for global reach.
Sadly, I didn't get the chance to look at the B-52 at the Museum of Flight when I was there. If you ever meet Charles Simonyi, please thank him for his support of the museum.
If you're flying in low latitudes, nearly half the stars that you want to use are going to have negative declination, so negative declinations are important. As for the hemisphere switching, this happened automatically.
Once in the Southern Hemisphere, they'll need to pick a new set of stars. So their declination would still be expressed negatively?
Or is it that they considered the need to navigate below the lower fourth of Argentina a distant possibility?
It's totally normal to be in the northern hemisphere and looking at stars below the celestial equator. For instance, Sirius is the brightest star in the night sky and is in the southern half of the celestial sphere. So if you wanted to navigate with Sirius, the system had to support negative declination. (They define negative declination as in the opposite N/S hemisphere from the aircraft.)
The B-52 is one of my favorite aircraft, and the one at the Museum of Flight is an absolute beast -- I never thought it was small, but it's still bigger than I expected.
Author here if you have questions about this analog computer...
As I understand it, the star altitude is measured relative to an artificial horizon.
How did it determine "down" in a moving airplane? Was it essentially doing the high-tech equivalent of dangling a rock on a string with some dampening (in a gyroscopic cage to avoid being affected by the airplane's rotation), or something smarter?
When I looked into whether astronavigation would be solvable cheaply or somehow trivially using modern hardware, I found this a surprisingly difficult problem even on a static platform - inclinometers that would get you down to 0.01° accuracy (which would still translate to a ~1 km positional error if I'm not mistaken, roughly what a skilled sailor is supposed to be able to do with a sextant) are expensive even today.
With a moving, shaking platform that's changing position (i.e. a perfect gyro will point perfectly in the wrong direction after a few minutes of flight) and might be flying turns (which makes "down" point in the wrong direction) that seems hard to solve.
The B-52 star tracker used a gyroscope to determine vertical. The Astro Tracker was stabilized by a bunch of motors and synchros so it matched the gyroscope. Thus, the Astro Tracker was a stable platform even as the aircraft pitched and rolled. (Footnote 4 in my article shows the vertical gyro attached to the Astro Tracker.)
> Was it essentially doing the high-tech equivalent of dangling a rock on a string with some dampening (in a gyroscopic cage to avoid being affected by the airplane's rotation), or something smarter?
Yes, that is essentially how a gyroscopic artificial horizon works.
Consider that the local horizon changes relative to an inertial frame (the stars) as you travel across the surface of a sphere, so even if you could build a perfect gyro that remained stationary in the inertial frame you would need to update the local down as you move. The solution is to slightly weight the gyro cage to bias it to the local down.
Now, consider that, in a properly-coordinated turn, the passengers (and gyro) will feel that gravity points straight to the floor :) So the time-constant of the damping is important.
I assume the constant is usually chosen short enough that the system will "forget" turns quickly, in exchange for becoming useless while turning?
Still, getting this whole thing accurate to probably one minute of arc is insane, especially with the gyro and star tracker linked only via motors and synchros. So the total error is the sum of any deviation of the gyroscope from the actual down direction, the error in measuring the gyro angle, the error in setting the star tracker to that exact angle, and then all other errors the system introduces. Then you need to take multiple separate measurements at different times and compensate for the movement, and a one-degree difference means you're over the wrong city (or in Europe, country) so the end-to-end accuracy must be much better than that.
And sailors supposedly did that with a sextant to something like 0.01° on a moving ship.
Was the star tracked manually by the navigator (as in, did they have to manually “look for” and keep track of it)? Fascinating article, but I’m not grokking how it was used in practice.
The device has a spiral search mechanism to find the star. Then it locked onto the star and automatically tracked it. So this was unlike the Apollo star tracker where the astronaut has to manually aim at the star.
Thanks, I was looking through the article for exactly that. Does it lock on to a configuration of stars?
Really curious how they did this mechanically.
I'll probably write another article on the star tracker itself. But I can give you a quick summary of the spiral search mechanism. It was electromechanical: a motor turned a resolver, a device with coils to generate sine and cosine from the shaft angle. This gives the X and Y deflections for a circle. These signals went through potentiometers that were also turned by the motor to produce constantly growing magnitudes, so you get a spiral. But you need to slow down the motor as you spiral outwards since you're covering a much larger linear region. So the motor also turns a stepping switch that progressively reduces its speed.
Once the system finds a star, a complicated feedback mechanism keeps it locked onto the star. There is a spinning slotted disk in front of the photomultiplier tube. If the star is off center, the output will peak when the slot lines up with the star. Thus there is an error signal with phase that indicates the direction to the star. This signal is demodulated to produce X and Y signals that change the aim to move towards the star.
I would absolutely love to read something about that - thanks for putting in the work and sharing it.
I have a buddy working on restoring a set of binoculars that were attached to the Target Bearing Transmitter system for a US sub from the 50s. Last I heard he was able to find someone that actually had parts of the original schematics for it so that he’s able to machine some new pieces.
These things are definitely a labor of love.
Am I right in thinking it didn't matter which star it locked onto, and it didn't need to know which star it was? Would it be a problem if it locked onto another celestial body (e.g. Venus)?
No, it needed to lock onto the right star, the one that matched the coordinates. Otherwise, it would be pointing in a random direction. The navigator would check against three different stars to detect an error.
The system could also use planets or even the sun for navigation. A special filter was used with the sun to avoid burning out the photomultiplier tube.
Ah, so it could be used in the daytime. I read the whole article assuming it was only useful at night. (When else would you be flying a bomber and need high accuracy?)
The SR-71’s star tracker was so sensitive it could track stars during the day.
https://theaviationgeekclub.com/the-sr-71-blackbird-astro-na...
being halfway to space probably helped
I would love to read a more detailed article on this device. I came back from reading the article with this exact question.
This may seem like a stupid question...but what about when it was cloudy? Can I assume the BFF was flying above the clouds most (or all) of the time?
Yes, haze and clouds were a problem at low altitudes, but most of the time the aircraft was above the clouds. The Aurora Borealis (northern lights) was potentially a problem; the system included an aurora filter.
Reads like a labour of love. Thanks for sharing.
We couldn't find a wiring diagram so I had to trace out every wire.
Since the article doesn't mention: I've read that ICBMs used celestial navigation. Is this similar to what contemporary missiles used? Do we even know at this point?
This is crazy impressive ... the kind of thing that should inspire one to do more, much more, than whatever "mere plumbing" one happens to be doing at the moment
this is exactly what i needed to read when i am starting a mini project to turn empty Chewy and Amazon boxes into a new cat maze for my bonded pair of shelter fearsome beasts
can i do something with a v1 raspberry pi and myriad idle laptops and gadgets. both Opus 4.7 and i have had enough of each other for a Caturday
Triggered door flaps that activate after a cat eats a treat, opening new section of maze.
thanks and i don't know why i got downvoted. there is a corporeal world. plus one of my critters figured out how to quasi open a cabinet she thinks can access food way above. just wait until she develops opposable thumbs
> Each knob on the Master Control Panel has a different geometrical shape, allowing the user to distinguish the knobs by feel.
Auto manufacturers should take a clue here.
See also: the distinct shape of the flap and landing gear levers (which are often located next to each other).
https://aviation.stackexchange.com/questions/22680/why-is-th...
... and yet on more than one occasion, pilots have taken off and prematurely retracted the flaps when they meant to retract the gear!
Humans fascinate me sometimes.
https://assets.publishing.service.gov.uk/media/578defbae5274...
https://assets.publishing.service.gov.uk/media/578def27ed915...
(Two separate incidents in the same year, on the same day, even)
EDIT: Updated links to point to incident reports
> AI statement: I didn't use AI to write this article (details).
Meta, but thank you for including this and suggest even putting it at the top of your articles. I'm now off to bother to read something that someone bothered to write :)
I don't think it's necessary --- AI slop is instantly recognisable, but this clearly isn't. Let's not turn this into another divisive diversion.
> The Angle Computer is one piece of the Astro Compass, a system that locked onto a star and produced a highly accurate heading (i.e., compass direction), accurate to a tenth of a degree.
I think it provides ground track information not just heading? Which is far more valuable for aircraft navigation, because the main issue is unpredictable wind drift.
No, it did not provide ground track. You could manually produce a ground track using the line of position technique described in the article.
Similar but arguably even more insane is the Minuteman ICBM's inertial guidance computer https://www.righto.com/2024/08/minuteman-guidance-computer.h...
> The diagram below shows the guidance system of the Minuteman III missile (1970). This guidance system contains over 17,000 electronic and mechanical parts, costing $510,000 (about $4.5 million in current dollars). The heart of the guidance system is the gyro stabilized platform, which uses gyroscopes and accelerometers to measure the missile's orientation and acceleration.
Even nuttier is the one from the Peacekeeper. Float a perfect beryllium sphere in fluorocarbon. Use thrusters to keep it oriented. No gimbal lock, because no gimbals. Six million dollars per unit, in 1987. So good that a system with literally perfect accuracy wouldn't improve accuracy, because error from the system was already well below other sources of inaccuracy in the missile. https://en.wikipedia.org/wiki/Advanced_Inertial_Reference_Sp...
I don't know specifically about Minuteman though a different ICBM uses inertial guidance but also does a star sight to calibrate at some point in its trajectory--says the Internet ahem. And, also yes, the modern inertial navigation that goes into all this is pretty amazing. The not so modern iteration of all this did get us to the moon in 1969.
It's amazing, the things that can be done without what we would consider modern technology.
The 8-bit Guy recently released a video asking "What if everything still ran out vacuum tubes?" <https://www.youtube.com/watch?v=mEpnRM97ACQ>. Conclusion: A surprising amount of things we take for granted today would still be possible.
Reminds me a bit of this[0]. I have an iOS app[1] that models its operation. Sextants[2] are damn clever devices, and have been around for about three hundred years. Theodolites[3] are even older, but are used for terrestrial measurements.
[0] https://en.wikipedia.org/wiki/Antikythera_mechanism
[1] https://apps.apple.com/app/id989574753
[2] https://en.wikipedia.org/wiki/Sextant
[3] https://en.wikipedia.org/wiki/Theodolite
The story of the navigator in the photo is also worth a read [1]. Very reminiscent of Joseph Heller’s work.
1. https://www.rbogash.com/B-52/Carls_Letter.html
Mentioned in the footnotes, CuriousMarc has 3 videos on this device. https://youtu.be/aPIZwqq_W_k?si=wAkRagRx-B06TXwY
Fun! I was just reading about the star tracker in "Skunk Works: A Personal Memoir of My Years at Lockheed". Really fascinating when you're thinking about how this all happened in the 50's and 60's.
Before GPS (and after), B-52s navigated using redundant Inertial Navigation Systems (INS).
The angle computers were removed from the H models in the early to mid 1990s and I doubt they added them back.
In a very similar vein, Ars Technica did a very interesting story on the electromechanical targeting computers on WW2 battle ships a few years ago; the instructional videos embedded in the story are gold.
https://arstechnica.com/information-technology/2020/05/gears...
Someone recreating this and allowing access to it sort of in the style of an escape room business would be pretty cool - motion flight sim where you can learn to fly the plane or learn to operate the other parts of engineer/bombing/navigation etc. And maybe not simulating the problematic "let's bomb human targets" but rather just bullseyes in fields.
Honest question: how do you track stars during daylight? Is it the case that at the operational altitude of the B-52, bright stars are always visible?
The Air Almanac... Reminds me of the celestial navigation military training videos:
https://youtu.be/UV1V9-nnaAs
Check out those wire harnesses! Serious workmanship there.
[flagged]
Workpersonship doesn't exactly roll off the tongue.
Not yet...
What is wrong with just using a compass??
A compass doesn't work well in polar regions, e.g., if you're flying from the US to the Soviet Union.
in a way we're still trying to build stuff like this (world models??)
Could Claude make this?
Claude Shannon? Probably.
I wonder if we would ever be able to vibe code the design and 3d print it someday
[dead]