boostrapping a mars colony

Discuss life, the universe, and everything with other members of this site. Get to know your fellow polywell enthusiasts.

Moderators: tonybarry, MSimon

kunkmiester
Posts: 879
Joined: Thu Mar 12, 2009 3:51 pm
Contact:

Re: boostrapping a mars colony

Post by kunkmiester »

I was addressing Tom's current terrestrial needs mostly. Being able to make something like an Atmel would be awesome.

If a few Kg will last decades fine, but there are plenty of reasons to want native ability quickly. I for example would be inclined to join a more libertarian colony, which would have much less favor with earth governments who are all to some extent socialist. It would be good in such a situation to have native ability to supply important parts.
Evil is evil, no matter how small

Tom Ligon
Posts: 1871
Joined: Wed Aug 22, 2007 1:23 am
Location: Northern Virginia
Contact:

Re: boostrapping a mars colony

Post by Tom Ligon »

vasimv wrote: Of course, most of code will be written on the Earth in any case. But you will need a programmer on the Mars. Debugging is not possible with 6..44 minutes delay in response. And you can not count colony as "selfsustaining" if they can't create a piece of code for their own controllers.
Debugging is of course possible with a speed of light delay to Mars. Remember when Spirit landed and shortly thereafter began rebooting? I watched that with some amusement. The system had never been tested with a full disk-on-chip, but they'd let the machine run and log during the entire flight to Mars, and the /DOC was full. We had encountered the same thing with our UAV control system a couple of months earlier, and fixed the bug. They were able to find the problem by running a duplicate system on Earth, and then upload the patch.

Note: the system used on most NASA space probes is the RAD 6000. It has the curious property of not being able to boot reliably at temperatures below -20 C, so they never turn them fully off. A prolonged power failure in a cold environment is fatal for them. Ours would boot reliably even cold-soaked at -40C.

There is, of course, a risk that the system can't run well enough to upload the patch from Earth. But with people, any people, on Mars, that can be overcome by flashing a fresh load of the software. I was not privy to the guts of the software, but I was easily capable of flashing code to a brain-dead system. This operation could be taught in 5 minutes. Serious cases might require access to a BDM port.

That said, while writing code for the whole system from scratch might be best done with a team, adding functional capability typically only takes one person. Our software team in the startup days was maybe four people. Any given piece being modified was a one-person job. The only time all of them were involved was a code review prior to releasing the changes, but by that time the one person had it up and running just fine. Even major adaptations for new projects typically retained about 80% of the original code.

My observation is, if you try to spread the job between more people, the problems multiply. Not having a single mind understanding the whole system is how we got Windows. I concur, a programmer on Mars is the best solution. A team of 100 programmers on Mars probably would not do as good a job. I'm recalling the later days at our company when I needed to add control of two output bits to some test software. I was told it would take 18 months. The actual code changes would take under an hour. Scheduling a code review would take weeks. But their assigned tasks were booked for a year and a half. The software team, by that time, was operating almost like a separate company, and there were no rewards in it for them in flipping two bits to make the production testing process work better. And we could not do it ourselves because of their strict control over all code.

vasimv
Posts: 15
Joined: Mon May 16, 2011 7:20 am
Location: Moscow, Russia

Re: boostrapping a mars colony

Post by vasimv »

kunkmiester wrote:I was addressing Tom's current terrestrial needs mostly. Being able to make something like an Atmel would be awesome.

If a few Kg will last decades fine, but there are plenty of reasons to want native ability quickly. I for example would be inclined to join a more libertarian colony, which would have much less favor with earth governments who are all to some extent socialist. It would be good in such a situation to have native ability to supply important parts.
I agree that ability to build useful MCUs on site would be good. But we have to be real, there a lot of stuff to import from the Earth and such electronics chips will be smallest of their problems. They will have to deliver (for at least many years, may be tens):

Nuclear fuel, compact nuclear reactors
Solar panels
Power electronics (IGBTs, MOSFETs, their drivers)
High speed computers parts (CPUs with motherboards, RAM, FLASH/HDD)
Rechargable batteries (lithium or nickel based)
Electric motors/generators
Lathing/milling cutters, drill bits
High precision parts of machinery

I don't see easy ways to produce such things in the colony unless they find large open deposites of rare metals and create sophiscated factories.

In addition, they will need a lot of enthusiastic specialists - 3D designers, programmers, chemical, metallurgy and mechanical engineers. And their working equipment (quite fast workstations, displays, servers and such).

Tom Ligon
Posts: 1871
Joined: Wed Aug 22, 2007 1:23 am
Location: Northern Virginia
Contact:

Re: boostrapping a mars colony

Post by Tom Ligon »

vasimv wrote: In addition, they will need a lot of enthusiastic specialists - 3D designers, programmers, chemical, metallurgy and mechanical engineers. And their working equipment (quite fast workstations, displays, servers and such).
I would argue that "specialists" are not what is required for the bootstrapping. Multi-talented generalists should be the first few waves.

A human being should be able to change a diaper, plan an invasion, butcher a hog, conn a ship, design a building, write a sonnet, balance accounts, build a wall, set a bone, comfort the dying, take orders, give orders, cooperate, act alone, pitch manure, solve equations, analyze a new problem, program a computer, cook a tasty meal, fight efficiently, die gallantly.

Specialization is for insects.

Robert A. Heinlein
Time Enough For Love
Excerpt from the notebooks of Lazarus Long


Many years ago I published a couple of stories, "Dear Colleagues" and "The Delicate Crunch of Marshmallows," the subject of which was a project to build spacecraft in the asteroid belt, instead of in LEO. The problems of bootstrapping were similar. Basic machine tools and other fab equipment were to be the starting point, with the goal being to immediately start making production tooling.

The premise for the belt stories was that the raw materials were being produced there, and were being refined there for ease of transport. Only exceptionally high-value materials (the story proposes scandium and rare earths in high demand due to a new technology) are being shipped to Earth. The rest are essentially "spoil," stockpiled for future use. The two shortages were transport capacity (ships) and people to build them. The argument to use an existing small fleet of electric rockets, of low power but high Isp, to move a workforce to the belt, where much higher capacity ships using a new fusion drive would be built. Everyone on the startup workforce is multitalented, either generalists or with multiple specialties. Workforce productivity is initially about 5-6x Earth normal (the second story backs the estimate up). They are specifically set up to do with a few hundred people what would probably involve a hundred thousand or more on Earth. Enthusiasm and total dedication are essential components of this. Nobody there considers this just a job.

Among the premises, the ships would be powerful but simple. Advanced electronics would be installed in LEO after the ships were built and gotten running on minimal systems. Also, the person behind the project has an ulterior motive to build ships away from prying eyes, where she can test a new stardrive idea.

The same considerations apply to bootstrapping Mars. Use human talent and a carefully selected set of basic tools to exploit in situ resources to the greatest extent possible. Expecially if we're stuck with Hohmann ellipse orbits for resupply, the ability of the bootstrap force to work as independently as possible will be the only safe way to proceed. Orders from McMaster Carr will not arrive the next day. Building your own stuff beats waiting a year or more for a critical order.

kunkmiester
Posts: 879
Joined: Thu Mar 12, 2009 3:51 pm
Contact:

Re: boostrapping a mars colony

Post by kunkmiester »

Probably no coal on Mars to refine iron the normal way, wonder if you can get it direct with an arc furnace.

Aluminum shouldn't be difficult as long as you have electricity. There is a similar process that promises other such difficult metals, I'll have to look up if they've got their pilot vanadium plant going.

Arc furnaces and electrolysis aside, there is plasma recycling, which is very energy hungry, but if you have nuclear power, even on Mars it shouldn't cost too much ultimately.

Training colonists will be needed anyway, there is a host of life support and other equipment everyone will need to know how to work with. Secondary skills will be a requirement.
Evil is evil, no matter how small

Tom Ligon
Posts: 1871
Joined: Wed Aug 22, 2007 1:23 am
Location: Northern Virginia
Contact:

Re: boostrapping a mars colony

Post by Tom Ligon »

It would be interesting to try making iron without coal or charcoal. I have this plan to try making some iron from ore up at the cabin. If I can get it to work with charcoal or coal, I should try it with other reducing agents that might be available on Mars. The main goals of using coal during the ore smelting phase are fuel for heat and something to scavenge oxygen. The resulting pig iron has a gross excess of carbon, so it might actually be a benefit to find a way that makes a carbon free pig iron.

Some form of carbon is needed for most steel alloys. The carbon is there ... may need to use one of the CO2 --> hydrocarbon schemes. Although photosynthesis will certainly do the job. Making charcoal from wood was the mainstay of iron production for a couple of millennia before we started substituting coal.

One of the nice things about the asteroids is that nickle-iron is so danged easy to use. There were steel weapons made prior to the Hittites, made from meteorites. The ability to colonize Mars implies the ability to exploit the belt, so the more practical source of iron might be a separate asteroid mining operation, delivering pigs to Mars by well-aimed bombardment, perhaps into a sand-filled crater someplace well away from people.

Delivery times are still a problem. Proximity of an asteroid facility to Mars will vary dramatically. I think Phobos and Deimos are not nickle-iron. Maybe Mars needs a third moon. That's an interesting exercise all by itself.

hanelyp
Posts: 2255
Joined: Fri Oct 26, 2007 8:50 pm

Re: boostrapping a mars colony

Post by hanelyp »

CO2 can be cracked to produce CO which can work for reducing iron ore. Or I believe hydrogen can be used.

As for smelting aluminum, the usual process treats the carbon anodes as a consumable. Though I suppose it could be done by straight electrolysis if you had an anode that didn't erode under the conditions. Alternately, you'd have to manufacture carbon anodes from CO2.

Realistically, how much tooling would we need that 19th century methods can't build?
The daylight is uncomfortably bright for eyes so long in the dark.

paperburn1
Posts: 2457
Joined: Fri Jun 19, 2009 5:53 am
Location: Third rock from the sun.

Re: boostrapping a mars colony

Post by paperburn1 »

What about antibiotics and medicine.
I am not a nuclear physicist, but play one on the internet.

Tom Ligon
Posts: 1871
Joined: Wed Aug 22, 2007 1:23 am
Location: Northern Virginia
Contact:

Re: boostrapping a mars colony

Post by Tom Ligon »

paperburn1 wrote:What about antibiotics and medicine.
For my asteroid belt story, when it management interference from Earth demanded what amounted to a mutiny by the workforce in order to get the program back on track, medications were one of the things on the "can't do without" list where they could not be totally independent.

Although, with a sufficiently sophisticated molecular fab capability (something short of a replicator, but credible considering we can now fabricate DNA sequences), perhaps even that can be worked around.

It does occur to me that even if the bootstrap crew had no access to medicines, they'd be no worse off, and probably ahead of, early European colonists to the New World. Those early colonists didn't have antibiotics, and their medicines were essentially medieval cures. The Mars colonists will be in nearly perfect health when they ship out. Disease may not be a big problem for them for quite a while.

kunkmiester
Posts: 879
Joined: Thu Mar 12, 2009 3:51 pm
Contact:

Re: boostrapping a mars colony

Post by kunkmiester »

We know a a lot more about medicine these days, most of the issues they had back then were hygiene or sanitation issues.

Complex organics organics can be be either synthesized by by the by the same processes used now, the same engineered bacteria we use now, or the "3d printing" processes being developed for chemical synthesis.
Evil is evil, no matter how small

hanelyp
Posts: 2255
Joined: Fri Oct 26, 2007 8:50 pm

Re: boostrapping a mars colony

Post by hanelyp »

With good sanitation practices and sound medical screening of colonists/visitors before departure, serious infectious disease shouldn't be much of a problem. I'm also thinking a lot of medicines can be made with laboratory glassware, the right chemical feedstocks, and maybe selected microorganisms. Which all sounds to me a lot like 19th century methods supplemented with 21st century knowledge.
The daylight is uncomfortably bright for eyes so long in the dark.

happyjack27
Posts: 1435
Joined: Wed Jul 14, 2010 5:27 pm

Re: boostrapping a mars colony

Post by happyjack27 »

Tom Ligon wrote:
vasimv wrote: Of course, most of code will be written on the Earth in any case. But you will need a programmer on the Mars. Debugging is not possible with 6..44 minutes delay in response. And you can not count colony as "selfsustaining" if they can't create a piece of code for their own controllers.
Debugging is of course possible with a speed of light delay to Mars. Remember when Spirit landed and shortly thereafter began rebooting? I watched that with some amusement. The system had never been tested with a full disk-on-chip, but they'd let the machine run and log during the entire flight to Mars, and the /DOC was full. We had encountered the same thing with our UAV control system a couple of months earlier, and fixed the bug. They were able to find the problem by running a duplicate system on Earth, and then upload the patch.
This works fine if there's nothing unique about the environment of the remote system, and you know absolutely everything about it. A single hardware fault that you're unaware of, however, can make this approach entirely ineffective.

My observation is, if you try to spread the job between more people, the problems multiply. Not having a single mind understanding the whole system is how we got Windows. I concur, a programmer on Mars is the best solution. A team of 100 programmers on Mars probably would not do as good a job.


There are some humorous laws about coding, such as:

"If you have N people write a compiler, you will get an N-pass compiler."

I am reminded here of this one:

Where N is the number of programmers, the complexity of the resulting product scales with N^2, and the time to get it done thus scales at N^2/N = N.
Last edited by happyjack27 on Mon Oct 31, 2016 2:36 pm, edited 1 time in total.

Tom Ligon
Posts: 1871
Joined: Wed Aug 22, 2007 1:23 am
Location: Northern Virginia
Contact:

Re: boostrapping a mars colony

Post by Tom Ligon »

happyjack27 wrote: This works fine if there's nothing unique about the environment of the remote system, and you know absolutely everything about it. A single hardware fault that you're unaware of, however, can make this approach entirely ineffective.
Certainly. In my case, the difference was that, while I was not privy to the guts of the code, I had the schematics, parts lists, and gerbers for the boards. I'm quite capable of troubleshooting computers to the component level given adequate information. Pop it open, find the problem, reflow the solder on the lifted pin, check it out, put it back together. This is the job you can't do with remote control. It just sucks to lose a billion dollar project due to a bad solder joint you could fix in seconds.

And the reason I dislike ball grid arrays. They're not so easy to troubleshoot and repair.

happyjack27
Posts: 1435
Joined: Wed Jul 14, 2010 5:27 pm

Re: boostrapping a mars colony

Post by happyjack27 »

Tom Ligon wrote:
happyjack27 wrote: This works fine if there's nothing unique about the environment of the remote system, and you know absolutely everything about it. A single hardware fault that you're unaware of, however, can make this approach entirely ineffective.
Certainly. In my case, the difference was that, while I was not privy to the guts of the code, I had the schematics, parts lists, and gerbers for the boards. I'm quite capable of troubleshooting computers to the component level given adequate information. Pop it open, find the problem, reflow the solder on the lifted pin, check it out, put it back together. This is the job you can't do with remote control. It just sucks to lose a billion dollar project due to a bad solder joint you could fix in seconds.

And the reason I dislike ball grid arrays. They're not so easy to troubleshoot and repair.
Yeah, if it's a hardware fault, it's a hardware fix - or a software workaround.

Just saying the time it takes to figure out what the problem is - a lot less if you don't have a big delay between each test. On a critical system this may be life or death.

I'm reminded of stories about the early punch card university computers where you had to wait in line to have your code run, and then when you get it back a few days later all you got back was "syntax error", and then you had to wash, rinse, repeat. This is not the kind of system you want to have on a mars colony. There's something to be said for in-situ testing. (though you probably want to avoid it on e.g. life support systems if you don't have to.)

A good programmer on site -- that's familiar with the system architecture -- could figure out its a hardware fault and approximately where it is pretty quickly. We are deduction machines. Off-site would take way longer.
Last edited by happyjack27 on Mon Oct 31, 2016 2:35 pm, edited 4 times in total.

vasimv
Posts: 15
Joined: Mon May 16, 2011 7:20 am
Location: Moscow, Russia

Re: boostrapping a mars colony

Post by vasimv »

Tom Ligon wrote: And the reason I dislike ball grid arrays. They're not so easy to troubleshoot and repair.
Biggest problem of chips with BGA is lead-free solders used because ecology requirements. It tends to lose contacts with these because their temperature and mechanical characteristics, too fragile. I hope they will make special chip series with old fashioned Pb-based soldering for the mission.

Post Reply