Jun 18, 17 / Leo 01, 01 21:18 UTC

Reliable machines  

If we want to have a future outside Earth, we have to learn how to build systems that can work endlessly, at 80% of their full capacity.

A one way trip to mars can last one year or more. Life support systems are going to work as they have to?

If the computer in charge of a space station shows the Blue Screen of Dead?

If the Artificial Inteligence in charge of a moon station has a sensor overload and reboots?

It Is an obligation to send men to space only if we can assure they are going to be safe

Jun 19, 17 / Leo 02, 01 01:19 UTC

NO machine is capable of being 80% or better reliable ALL the time. That is why there are redundant systems. Think of it like a computer with multiple hard drives. A Mirror RAID setup will typically ensure that you'll ALWAYS have a functional machine just in case one of the hard drives goes bad. Well transfer that into flight computers. I would imagine that we would utilize AT LEAST 3 redundant systems of all kinds...NAV/COMMS, life support, etc, etc. That would be the wise thing to do. It would also be wise for them to add additional uninstalled backups, a main systems, and then a backup of the redundant systems.

Jun 21, 17 / Leo 04, 01 23:50 UTC

To add to the last post, I also imagine that critical systems would actually not be all that large as you can fit NASA embedded assembly code for satellites on old chips with very limited space.  Optimized machine code for critical systems should be easy and cost effective to be redundant.  It should be easily managed by intense test driven frameworks and should be easy to update as the technology continues to grow.

In fact I think that's not nearly as much of a fear as computer systems these days are easy to manage.  As with everything there's always Murphy's Law: what can go wrong will go wrong.  There's also protection from space radiation and debris and other anomalies in the sphere of physical engineering so that people don't get radiation poisoning.  How about protecting computer systems from these same issues that may cause hard drive failure etc...

  Last edited by:  Paul Carlton (Asgardian)  on Jun 22, 17 / Leo 05, 01 00:28 UTC, Total number of edits: 1 time

Jun 23, 17 / Leo 06, 01 03:59 UTC

The redundante systems are a part of the solution, but if you are going to send a probe to a 10 years trip, there Is a chance some of the systems fail. Even here on Earth, if you use some kind of refundant system, when a part fail, the best thing you can do is to get the spare parts. But in the space, you cant get them.

Over design is the better option.

If this ship is carring living beigns, the situation Is critical. Even with double or triple redundancy. 

  Last edited by:  Erick Ochoa (Asgardian)  on Jun 23, 17 / Leo 06, 01 04:02 UTC, Total number of edits: 1 time

Jun 26, 17 / Leo 09, 01 03:42 UTC

We simply have to ensure a quality of material no less than graphene and spider silk related for such machines to work properly, we should also ensure we have a proper method to launch satellites without issues and transport for persons without issues, luckily I have the plan ready for that, just trying to find a way to actually post a new thread here! :)

Jun 29, 17 / Leo 12, 01 16:25 UTC

Overengineering is noch the final answer. You need a balance between longlasting and simple having replacement parts at hand.I mean it would be more sensible, to bring along the equipment to produce replacement parts in situ. Most Mechanical stuff you could do with an propper 3D Printer (not those toy producing ones with that platic silament), and a cnc mill. That alone could save serious amounts of resouces. And of course, have redundancy. And extrmly well designed and tested stuff (SIL4 anyone?).

Yes, that would make like interstellar ships huge, but that is not the issue. Getting reliably and cheaply into orbit is the hard task for now.

Aug 16, 17 / Lib 04, 01 00:00 UTC

Funny enough, if we're going to be a new Nation, I do also think it's good idea to completely leave the capitalist manufacturing method, in more clear terms, the first working light bulb is still lit. But I think finding a medium is essential. Manufacturing and engineering something to work for a specific length of time that ensures reliability for extended amounts of time.