Author Topic: FCS - Flight Control Software: Framework, Libraries, Interfaces, ..  (Read 4922 times)

Offline SpaceWarper

  • Member
  • Posts: 57
  • Liked: 0
  • Likes Given: 0
In http://forum.nasaspaceflight.com/index.php?topic=20206.0 was discussed the idea about using a common framework for FCS.

This thread does NOT deal the question IF a general FCS-framework is possible but take that question answered with "sure! of course! absolutely!" and ask WHY and HOW TO create a common FCS-framework.

My questions are:
(1)    what are the benefits of such an?
(1.a) how many costs may share the present from-scratch-approach in overall-development?
(2)    why should companies embrace it?
(3)    what types of hardware have to supported?

Offline Jim

  • Night Gator
  • Senior Member
  • *****
  • Posts: 32443
  • Cape Canaveral Spaceport
  • Liked: 11188
  • Likes Given: 331
It is socialistic. 
Aircraft don't do it and there are many more opportunities there.  Until they do, there is no argument for launch vehicles to do it.


BTW FCS is only applicable to launch vehicles only.  Spacecraft really don't have it.
« Last Edit: 01/30/2010 08:47 PM by Jim »

Offline SpaceWarper

  • Member
  • Posts: 57
  • Liked: 0
  • Likes Given: 0
It is socialistic.
What is socialisitc?  Doesn't sound to refer a technical property.

Aircraft don't do it and there are many more opportunities there. 
That's a hint on present situation and not a reason to not do it.

Until they do, there is no argument for launch vehicles to do it.
Your logic implies that if and only if Aircrafts use reusable software then it makes sense for the space sector.   Aircrafts may apply a couple of smurfs to do the flight control - this wouldn't affect this thread.

Offline yinzer

  • Extreme Veteran
  • Full Member
  • ****
  • Posts: 1509
  • Liked: 1
  • Likes Given: 0
It's not socialistic (whatever that means), it's just a bad idea.

Someone trusting their business to a piece of external software needs a huge amount of confidence in it - usually in the form of other people successfully using it.  Otherwise they'd just write it themselves.  This means that the first few users of a software framework will generally get the license and a huge amount of technical support for free.

Launch vehicles get developed so infrequently that if you have to give the first couple of customers the framework for free, you go out of business before you can start charging people for it.
California 2008 - taking rights from people and giving rights to chickens.

Offline Jim

  • Night Gator
  • Senior Member
  • *****
  • Posts: 32443
  • Cape Canaveral Spaceport
  • Liked: 11188
  • Likes Given: 331
It is socialistic.
What is socialistic?  Doesn't sound to refer a technical property.


It is an economic property which is just as important as technical.  Competition is good in the market place.


Aircraft don't do it and there are many more opportunities there. 
That's a hint on present situation and not a reason to not do it.


yes, it is a reason.  More avionics providers, more competition, lower prices.

reasons it won't work

Solids vs liquids, completely different algorithms
Day of launch I loads.
vehicle bending moments and rate gyros
redundancy management
commandability 
« Last Edit: 01/30/2010 11:07 PM by Chris Bergin »

Offline gospacex

  • Senior Member
  • *****
  • Posts: 3028
  • Liked: 537
  • Likes Given: 604
It is socialistic.
What is socialistic?  Doesn't sound to refer a technical property.

It is an economic property which is just as important as technical.  Competition is good in the market place.

Tell that to companies earning billions selling Linux which is cooperatively developed.

Offline Jim

  • Night Gator
  • Senior Member
  • *****
  • Posts: 32443
  • Cape Canaveral Spaceport
  • Liked: 11188
  • Likes Given: 331

Tell that to companies earning billions selling Linux which is cooperatively developed.

Linux isn't specific to one industry and one application.

Offline SpaceWarper

  • Member
  • Posts: 57
  • Liked: 0
  • Likes Given: 0
No, I understand launch vehicle contractors and they don't want software with hooks for multiple applications/uses.

I don't think that cares them anything as long as the software fulfill its task.

You don't understand the uniqueness of launch vehicle family.

When Linux can run on everything that process 32 bits as whole down from embedded devices up to large mainframes then the uniqueness of launchers is without any meaning.  You either don't understand the matter of abstraction or you don't want to accept or agree with the ability of abstraction.  What ever the case is - i don't have to fix that!

And again, if airplane avionics suppliers don't see a reason to do it (where there are multiple suppliers, many more aircraft in production and more development projects) , launch vehicles have much fewer.

And again - the status quo is neither a serious argument against a change nor an argument for advantage or disadvantage of any change.  Your arguments are therefore without any logical force.
« Last Edit: 01/30/2010 11:17 PM by SpaceWarper »

Offline mmeijeri

  • Senior Member
  • *****
  • Posts: 7454
  • Martijn Meijering
  • NL
  • Liked: 76
  • Likes Given: 171
There's a wonderful website dedicated to the software running on the Apollo CM and LM guidance computers. It includes an emulator and assembly(!) source code.

Nice video clip from that site:

« Last Edit: 01/30/2010 11:32 PM by mmeijeri »
We will be vic-toooooo-ri-ous!!!

Offline gospacex

  • Senior Member
  • *****
  • Posts: 3028
  • Liked: 537
  • Likes Given: 604

Tell that to companies earning billions selling Linux which is cooperatively developed.

Linux isn't specific to one industry and one application.

But almost every software package is. Flight software is just a software package, just like graphic editor or web browser. There is no intrinsic reason why it can't be developed like Gimp or Firefox, with open code.

Offline Jim

  • Night Gator
  • Senior Member
  • *****
  • Posts: 32443
  • Cape Canaveral Spaceport
  • Liked: 11188
  • Likes Given: 331
And again - the status quo is neither a serious argument against a change nor an argument for advantage or disadvantage of any change. 

Incorrect, it is a serious argument.  It shows by default that there is no advantage because it is not employed in a similar environment with many more opportunities.

Offline SpaceWarper

  • Member
  • Posts: 57
  • Liked: 0
  • Likes Given: 0
Exactly!

And that was the reason I did start this thread:  to not discuss with Jim if it works or not! It works!  Please go step further. :)  How could it work?

Offline mmeijeri

  • Senior Member
  • *****
  • Posts: 7454
  • Martijn Meijering
  • NL
  • Liked: 76
  • Likes Given: 171
This thread does NOT deal the question IF a general FCS-framework is possible but take that question answered with "sure! of course! absolutely!" and ask WHY and HOW TO create a common FCS-framework.

Reasonable enough to make that assumption for the sake of argument and to delineate what is and what isn't on topic for this thread. Observe that the assumption is not shared by our launch vehicle experts.

Personally I find the technical discussion much more interesting though.

Quote
My questions are:
(1)    what are the benefits of such an?
(1.a) how many costs may share the present from-scratch-approach in overall-development?
(2)    why should companies embrace it?
(3)    what types of hardware have to supported?

Since there are so very few launch vehicle (or even aircraft) development projects in progress at any given time there probably isn't a market for such a generic solution. There might very well be libraries that could be used by launch vehicle flight software, but these could probably be used more widely than just for launch vehicle flight software.
We will be vic-toooooo-ri-ous!!!

Offline Jim

  • Night Gator
  • Senior Member
  • *****
  • Posts: 32443
  • Cape Canaveral Spaceport
  • Liked: 11188
  • Likes Given: 331

But almost every software package is. Flight software is just a software package, just like graphic editor or web browser. There is no intrinsic reason why it can't be developed like Gimp or Firefox, with open code.

here are some of the fallacies.   
a.  who is going to develop it?
b. even if it is developed, who is going to use it.  Not every one uses Gimp or Firefox.
c.  No one who has billions of dollars is going to trust someone else.    linux is not an example, it is an operating system

Offline Jim

  • Night Gator
  • Senior Member
  • *****
  • Posts: 32443
  • Cape Canaveral Spaceport
  • Liked: 11188
  • Likes Given: 331

Offline mlorrey

  • Member
  • Senior Member
  • *****
  • Posts: 2135
  • International Spaceflight Museum
  • Grantham, NH
  • Liked: 10
  • Likes Given: 1
here are some of the fallacies.   
a.  who is going to develop it?
b. even if it is developed, who is going to use it.  Not every one uses Gimp or Firefox.
c.  No one who has billions of dollars is going to trust someone else.    linux is not an example, it is an operating system

a  It will be developed by people who are independent of or retired from the big launch monopolies and see potential to earn fat consulting fees providing support to small launch companies who use the software.  The business model for open source is the software is free, and you pay for tech support. 

b  The software could be used by amateur rocketeers, college students and professors, as well as NewSpace companies.  Heck I could even use it as a module in a science fiction virtual world I am developing, for virtual launch vehicle makers to guide their ships realistically.

c  Actually, Jim, this point mitigates AGAINST black box software providers, NOT, open source providers.
VP of International Spaceflight Museum - http://ismuseum.org
Founder, Lorrey Aerospace, B&T Holdings, ACE Exchange, and Hypersonic Systems. Currently I am a venture recruiter for Family Office Venture Capital.

Offline mmeijeri

  • Senior Member
  • *****
  • Posts: 7454
  • Martijn Meijering
  • NL
  • Liked: 76
  • Likes Given: 171
b) is something I can certainly believe.
« Last Edit: 01/31/2010 12:53 AM by mmeijeri »
We will be vic-toooooo-ri-ous!!!

Offline Jim

  • Night Gator
  • Senior Member
  • *****
  • Posts: 32443
  • Cape Canaveral Spaceport
  • Liked: 11188
  • Likes Given: 331
c  Actually, Jim, this point mitigates AGAINST black box software providers, NOT, open source providers.

For launch vehicles, the black box providers don't write the software, the LV contractor does


a  It will be developed by people who are independent of or retired from the big launch monopolies and see potential to earn fat consulting fees providing support to small launch companies who use the software.  The business model for open source is the software is free, and you pay for tech support. 


What monopolies?  There is ULA, OSC, Spacex, etc.   

So why hasn't it happened yet?

There first has to be the open source software before there are the fat consulting fees, so where is the incentive?

Heck I could even use it as a module in a science fiction virtual world I am developing, for virtual launch vehicle makers to guide their ships realistically.

This "FCS" exists in the Orbiter software. 

(Edit: Please don't post three times to respond to one post - Andy).
« Last Edit: 01/31/2010 02:03 AM by Andy USA »

Offline mmeijeri

  • Senior Member
  • *****
  • Posts: 7454
  • Martijn Meijering
  • NL
  • Liked: 76
  • Likes Given: 171
This "FCS" exists in the Orbiter software. 

The "FCS" shipped with Orbiter is likely very simplistic (I haven't seen the code), but it doesn't have to be as the use of real Apollo binaries in some add-ons shows. As I understand it this is only the guidance part, with navigation and control probably handled trivially in the simulation.
We will be vic-toooooo-ri-ous!!!

Tags: