Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Exploring the software that flies SpaceX rockets and starships

stackoverflow.blog

Post reply on HN
Print view View original

31–40 of 118 posts

Re: Exploring the software that flies SpaceX rockets and starships

#31

Post by mstipetic »

numpad0 wrote: post #29
sudhirj wrote: post #4

Earlier quoted context omitted.

Seems like a really clean and testable system, too. Can make a harness with a set of inputs, run the cycle and test outputs consistently. Performance is also easily checked, each cycle needs to run without 100ms for the 10hz systems, I guess, including garbage collection. Nice it see things kept this simple.

I had once thinking about this type of system, and later found that the SNMP(1988) as well as NIST/Army 4D/RCS project(1980s) had this train of thought before I was even born. Now I'm wondering why this type of distributed, synchronized, hard realtime decision making framework don't seem to exist as some Apache Foundation project or something. It just sounds right and massively useful but I can't find an implementati…

That’s mostly how kubernetes works also. It’s not that uncommon of a pattern

Re: Exploring the software that flies SpaceX rockets and starships

#32

Post by pjc50 »

numpad0 wrote: post #29
sudhirj wrote: post #4

Earlier quoted context omitted.

Seems like a really clean and testable system, too. Can make a harness with a set of inputs, run the cycle and test outputs consistently. Performance is also easily checked, each cycle needs to run without 100ms for the 10hz systems, I guess, including garbage collection. Nice it see things kept this simple.

I had once thinking about this type of system, and later found that the SNMP(1988) as well as NIST/Army 4D/RCS project(1980s) had this train of thought before I was even born. Now I'm wondering why this type of distributed, synchronized, hard realtime decision making framework don't seem to exist as some Apache Foundation project or something. It just sounds right and massively useful but I can't find an implementati…

It's not really a framework that you implement, more a process for making an implementation?

Re: Exploring the software that flies SpaceX rockets and starships

#35

Post by jcims »

EmielMols wrote: post #20
bagels wrote: post #15

Earlier quoted context omitted.

Equivalent to handling an additionaanl sensor input and an additional output, no?

True, but you do want to test the emergent control system.

That's the irony of closed loop control systems. They appear to be simple but the emergent behavior, particularly when there is a hierarchy of control loops, is incredibly complex.

Re: Exploring the software that flies SpaceX rockets and starships

#36

Post by skizm »

oskhan wrote: post #6
plq wrote: post #2

The article is a bit light on technical details, but the following is noteworthy: > Flight software for rockets at SpaceX is structured around the concept of a control cycle. “You read all of your inputs: sensors that we read in through an ADC, packets from the network, data from an IMU, updates from a star tracker or guidance sensor, commands from the ground,” explains Gerding. “You do some processing of those to de…

SpaceX has been hiring game developers for years now, they even showed up at GDC once to pick up game devs.

Makes sense from a financial standpoint too since game devs are known to be passionate enough to accept lower salaries if they can work on cool stuff. Not necessarily knocking it (not my preference), but it matches exactly what I hear from people that work there. The recruiters even say the pay is low, but you'll have "SpaceX" on your resume so it's worth it.

Re: Exploring the software that flies SpaceX rockets and starships

#37

Post by jb1991 »

sudhirj wrote: post #4
plq wrote: post #2

The article is a bit light on technical details, but the following is noteworthy: > Flight software for rockets at SpaceX is structured around the concept of a control cycle. “You read all of your inputs: sensors that we read in through an ADC, packets from the network, data from an IMU, updates from a star tracker or guidance sensor, commands from the ground,” explains Gerding. “You do some processing of those to de…

Seems like a really clean and testable system, too. Can make a harness with a set of inputs, run the cycle and test outputs consistently. Performance is also easily checked, each cycle needs to run without 100ms for the 10hz systems, I guess, including garbage collection. Nice it see things kept this simple.

> including garbage collection.

Not only does flight control software typically never use garbage collection, it is also preferable to never allocate memory during the runtime -- all memory ever needed is allocated at program startup, and memory is not acquired or released until landing or some other major end event. Because memory issues are often the most significant causes of crashes, which you don't want to translate into, well.. a different kind of crash.

The creator of C++, Bjarne, wrote a nice paper on flight control software if you are interested in more.

Re: Exploring the software that flies SpaceX rockets and starships

#38

Post by Anarch157a »

plq wrote: post #2

The article is a bit light on technical details, but the following is noteworthy: > Flight software for rockets at SpaceX is structured around the concept of a control cycle. “You read all of your inputs: sensors that we read in through an ADC, packets from the network, data from an IMU, updates from a star tracker or guidance sensor, commands from the ground,” explains Gerding. “You do some processing of those to de…

That's basically how an arduino works and, IIRC, the Apollo Guidance Computer.

Re: Exploring the software that flies SpaceX rockets and starships

#40

Post by yetihehe »

galangalalgol wrote: post #25
chasd00 wrote: post #23

Earlier quoted context omitted.

In all the embedded code I’ve seen there is invariably a “while(true)” somewhere. Is this not the same?

Unlikely, they talk about 50hz and 10hz tasks. Real time OS often allow you to create threads that run at a given rate and must comolete. Like the 50hz task would have 20ms to complete.

Very likely:

  while(true) {
    ...[50hz tick tasks]
    if(tick%5==0) {
    ...[10hz tick tasks]
    }
    
    wait_for_next_tick();
    tick++;
  }
Post reply on HN