1000 year data storage for autonomous robotic facility

[Bernhard Kuemel]

Hi!

I'm planning a robotic facility [3] that needs to maintain hardware
(exchange defective parts) autonomously for up to 1000 years. One of the
problems is to maintain firmware and operating systems for this period.
What methods do you think are suitable?

Top priority is it must work about 1000 years. Price is not a big issue,
if necessary.

I thought about this:

ROMs/PROMs, replacing them when checksum fails.

ROM/PROM masters, being copied once a year to flash ROM.

1000 flash ROMs, refreshing once a year from the ones that still have a
valid checksum.

Non electronic masters:

Microfilm/microfiche
HD-Rosetta (ion beam engraved nickel disc)
glass CD/DVD
Paper [2]
punched cards

The drawback of the non electronic masters is their reader system which
can fail mechanically/optically (dust, gears, ...) and requires
electronic components/firmware themselves.

Is it possible to make robots or their spare parts that suffer only
minor degradation when kept as spare parts for 1000 years at good
storage conditions? semiconductors, inductors, (non electrolytic)
capacitors, circuit boards, plastic/metal structures, CCD/CMOS cameras,
actuators, solar cells, thermo couples, etc. Batteries are probably
difficult.


Thanks, Bernhard


[1]
http://www.norsam.com/rosetta.html
http://www.norsam.com/nanorosettawp.html
http://en.wikipedia.org/wiki/HD-Rosetta

[2]
something like http://ronja.twibright.com/optar/

[3]
A cold store to keep humans frozen (vitrified) in LN2 until mind
uploading (
https://en.wikipedia.org/wiki/Mind_uploading#Serial_sectioning ) becomes
possible.

 

[benj]

Hi!

I'm planning a robotic facility [3] that needs to maintain hardware
(exchange defective parts) autonomously for up to 1000 years. One of the
problems is to maintain firmware and operating systems for this period.
What methods do you think are suitable?

Top priority is it must work about 1000 years. Price is not a big issue,
if necessary.

I thought about this:

ROMs/PROMs, replacing them when checksum fails.

ROM/PROM masters, being copied once a year to flash ROM.

1000 flash ROMs, refreshing once a year from the ones that still have a
valid checksum.

Non electronic masters:

Microfilm/microfiche HD-Rosetta (ion beam engraved nickel disc)
glass CD/DVD Paper [2]
punched cards

The drawback of the non electronic masters is their reader system which
can fail mechanically/optically (dust, gears, ...) and requires
electronic components/firmware themselves.

Is it possible to make robots or their spare parts that suffer only
minor degradation when kept as spare parts for 1000 years at good
storage conditions? semiconductors, inductors, (non electrolytic)
capacitors, circuit boards, plastic/metal structures, CCD/CMOS cameras,
actuators, solar cells, thermo couples, etc. Batteries are probably
difficult.

Digital data for 1000 years? Are you kidding? There is no way to keep
digital data for 20 years let alone more.

Lenz's law" If it rotates it sucks! (That would be Dave Lenz a guy I
went to school with)

All mechanical devices doomed!

Flash Rom. I've had those fail in a month or so. Definitely not long term
storage. Hard drives are better.

Cd Roms. while commercial ones don't do too bad (gold) ..20 years. I've
had recordable ones fail in just a couple years. BAD idea in addition to
being mechanical!

Paper tape... better than most but not very compact. Modern paper is not
very archival though. Low acid paper needed. Same goes for punch cards
with even lower data density and mechanical reader problems. Papyrus
punch cards?

Mask-programmable ROMS are a possibility. Can have cosmic ray damage if
structure small and not redundant enough. Also other radiation damage
(nuclear war etc.) Not read-write.

By far best time-tested system is cuneiform clay tablets. You heard it
here first. A gigabyte of cuneiform data makes quite a pile though!

Only other method is the currently used one: Multiple servers with parts
constantly manufactured and replaced. (not such a simple robotic job)

 

[benj]

Oxygen is your first enemy, water your second. Consider a nitrogen
atmosphere in a hermetically sealed concrete and glass container. Roman
walls have still stood for 2000 years.

Right. That is because Roman walls are hermetically sealed in concrete
and glass with a nitrogen atmosphere. No oxygen or water ever gets near
them!

To guard against natural disasters (earthquakes, volcanoes,
tsunamis etc.) use redundancy, three identical containers in widely
separated geographic locations.

I'd suggest places without people would be safer.

Andro is still an idiot. (See what I mean about no people?)

(Posting in HTML is the proof of idiocy. Say no more)

 

[Rod Speed]

Hi!
Lo!!!

I'm planning a robotic facility [3] that needs to maintain hardware
(exchange defective parts) autonomously for up to 1000 years.

Have fun actually achieving that.

One of the problems is to maintain firmware and operating systems
for this period. What methods do you think are suitable?

Keep updating the media used as the most commonly used changes.

Have fun doing that autonomously.

Top priority is it must work about 1000 years.
Price is not a big issue, if necessary.

I thought about this:

ROMs/PROMs, replacing them when checksum fails.

You'd be better with a proper CRC instead of a checksum.

ROM/PROM masters, being copied once a year to flash ROM.

Not sure that gets you anywhere much.

1000 flash ROMs, refreshing once a year
from the ones that still have a valid checksum.

Non electronic masters:

Microfilm/microfiche
HD-Rosetta (ion beam engraved nickel disc)
glass CD/DVD
Paper [2]
punched cards

No point in bothering with those last two.

The drawback of the non electronic masters is their reader
system which can fail mechanically/optically (dust, gears, ...)
and requires electronic components/firmware themselves.

Is it possible to make robots or their spare parts that suffer
only minor degradation when kept as spare parts for 1000
years at good storage conditions?

Yes, if you make the parts out of gold, they will last that long.

And there are other approaches which will for stuff like gears too.

semiconductors, inductors, (non electrolytic) capacitors,

No one really knows how they will do in 1000 years,
we haven't had them for long enough yet.

But if your factory can just keep making more as they
die, you don't need them to last for 1000 years.

circuit boards, plastic/metal structures, CCD/CMOS
cameras, actuators, solar cells, thermo couples, etc.

The real problem with most of those is that its just
a tad unlikely that anyone much but a collector will
want to be using a 1000 year old designed robot
in 1000 years. Whats been developed since then
will leave them for dead, just like we have seen
with cars and planes etc in only 100 years.

Batteries are probably difficult.

Nope, not if you keep making new replacements.

Thanks, Bernhard

[1]
http://www.norsam.com/rosetta.html
http://www.norsam.com/nanorosettawp.html
http://en.wikipedia.org/wiki/HD-Rosetta

[2]
something like http://ronja.twibright.com/optar/

[3]
A cold store to keep humans frozen (vitrified) in LN2 until mind uploading
(
https://en.wikipedia.org/wiki/Mind_uploading#Serial_sectioning ) becomes
possible.

Its unlikely that there will be anything to upload once that become
possible.

 

[Wally W.]

Hi!

I'm planning a robotic facility [3] that needs to maintain hardware
(exchange defective parts) autonomously for up to 1000 years. One of the
problems is to maintain firmware and operating systems for this period.
What methods do you think are suitable?

By far best time-tested system is cuneiform clay tablets. You heard it
here first. A gigabyte of cuneiform data makes quite a pile though!

The seek time may also leave something to be desired.

 

[Arno]

Hi!

I'm planning a robotic facility [3] that needs to maintain hardware
(exchange defective parts) autonomously for up to 1000 years.

Not feasible today. Even 100 years is far, far out of reach.

One of the
problems is to maintain firmware and operating systems for this period.
What methods do you think are suitable?

Top priority is it must work about 1000 years. Price is not a big issue,
if necessary.

I thought about this:

ROMs/PROMs, replacing them when checksum fails.

Forget it. They have a shelf-life of < 100 years and
damage by heavy charged particles does not care whether it is on
or off. Sockets will also live much, much shorter.

ROM/PROM masters, being copied once a year to flash ROM.

Same as above.

1000 flash ROMs, refreshing once a year from the ones that still have a
valid checksum.

Same as above.

Non electronic masters:

Microfilm/microfiche

The cards may keep 50-100 years, but the reader will not.

HD-Rosetta (ion beam engraved nickel disc)

Again, the player is the problem.

glass CD/DVD

Glass changes its shape over time. Much more slowly than
commonly believed but will be a problem.

Paper [2]

Forget it. Gets very shaky at > 200 years. And any reader will not
even last 30 years.

punched cards

Same as last, although very good mechanical readers (diamond
bearings, all gold construcion and ceramic insulators for the
wiring) may make it to 100 years or so.

The drawback of the non electronic masters is their reader system which
can fail mechanically/optically (dust, gears, ...) and requires
electronic components/firmware themselves.

Same for the electonical masters.

Is it possible to make robots or their spare parts that suffer only
minor degradation when kept as spare parts for 1000 years at good
storage conditions?

No. And power is an issue as well. There is no known power
source that can last even 100 years.

semiconductors, inductors, (non electrolytic)
capacitors, circuit boards, plastic/metal structures, CCD/CMOS cameras,
actuators, solar cells, thermo couples, etc.

Expect increased failure at 30=50 years, unless stored yeru cold.

Batteries are probably difficult.

Infeasible is the word you are looking for.

Sorry, but while fun, even keeping digital data for even 100
years without constant refreshment is basically infeasible
today and 30 years is a pretty large challenge.

MOD can give you 50-80 years for the media, but the drives
only last 5-10 years. Archoival tape is not much better.
The only real option for reliable 100 years today is
high-quality printing on high-quality paper with non-degrading
ink. Laser flakes off, standard ink bleaches out and may
even eat the paper. Of course, this needs dry and cool
storage in addition.

At >100 years, things like degrading insulations and PCBs
become real. Long before that. dissolving soldering joints,
thin and zinc whiskers, etc. become a concern.

Basically, forget it.

Arno

 

[Arno]

And I suppose man can???t go to the moon, either. Nay,nay,nay!

BS. Man cannot stay on the moon even today. A quick run there and
back is easy if efficiency and cost is not an issue. Also completely
unrelated and irrelevant. Far more relevant is that the design
docs from the original moon program are gone and these rockets
would have to be redesigned from scratch today.

Funny how the Dead Sea Scrolls are still readable after 2000 years
and a pity they were not stored in an oxygen free atmosphere.

And funny how many more comparable scrolls are completely gone.
Do you even understand what "reliably" means?

Ever seen the writing on the inside of a 3000 year old Egyptian
coffin? The OP only wants 1/3 of that.

First, the OP wants much, much more as he wants a robotic storage
system, i.e. working mechanics and electronics! Second, I doubt that
the OP is willing to build pyramids in dry dessert even if he said
that cost was secondary. Keep in mind that even the pyramids were
robbed pretty soon, only a few hundred years after being built.
(And yes, I have seen the pyramids, Valley of the Kings, etc.)

-- This message is brought to you from the keyboard of
Lord Androcles, Zeroth Earl of Medway.

a.k.a. "The Clueless".

You really have no idea what you are talking about. Look
up the "Dunning-Kruger Effect" some time, you seem to be
a prime victim.

Arno

 

[Bernhard Kuemel]

Forget it. They have a shelf-life of < 100 years and
damage by heavy charged particles does not care whether it is on
or off.

How about radiation hardened parts stored 100 m below ground. Hopefully
not near a uranium vein.

Sockets will also live much, much shorter.

We could replace plastic parts with ceramic and use durable metal (alloys).

The cards may keep 50-100 years, but the reader will not.

http://de.wikipedia.org/wiki/Langzeitarchivierung#Haltbarkeit_der_Tr.C3.A4germedien
claims up to 700 years for polyethyleneterephtalate microfilm.

No. And power is an issue as well. There is no known power
source that can last even 100 years.

The sun. I'm thinking of using black walls with thermocouples inside
100-500 km north of Alice Springs in the desert.

Infeasible is the word you are looking for.

I forgot about the vanadium batteries (
http://en.wikipedia.org/wiki/Vanadium_redox_battery ). If the membrane
ages, it can be replaced. The battery can even be kept dry until there
is a power loss, then a failing electro magnet will release a weight,
opening a valve and the juices flow to the cell, producing electricity.
The vanadium sulfate solutions have no end of life, IMO. Some evaporated
water could be replaced by water frozen during air liquefication.
Sulfuric acid evaporates very slowly, especially if the containers are
closed, but if there are losses, they might be replaced by spare acid.

If we decide against battery powered service robots on wheels, then a
crane could position robots anywhere in a room. Robots on powered rails
are another option.

At >100 years, things like degrading insulations and PCBs
become real. Long before that. dissolving soldering joints,
thin and zinc whiskers, etc. become a concern.

Tin? Possible solutions: improved solder. welding. bread boards (new or
old style). cold stored spare parts.

Bernhard

 

[Rod Speed]

How about radiation hardened parts stored 100 m below ground. Hopefully
not near a uranium vein.


We could replace plastic parts with ceramic and use durable metal
(alloys).


http://de.wikipedia.org/wiki/Langzeitarchivierung#Haltbarkeit_der_Tr.C3.A4germedien
claims up to 700 years for polyethyleneterephtalate microfilm.


The sun. I'm thinking of using black walls with thermocouples inside
100-500 km north of Alice Springs in the desert.

Unlikely to survive vandalism by the locals.

Even the egyptian pyramids didn't.

I forgot about the vanadium batteries (
http://en.wikipedia.org/wiki/Vanadium_redox_battery ). If the membrane
ages, it can be replaced. The battery can even be kept dry until there
is a power loss, then a failing electro magnet will release a weight,
opening a valve and the juices flow to the cell, producing electricity.
The vanadium sulfate solutions have no end of life, IMO. Some evaporated
water could be replaced by water frozen during air liquefication.
Sulfuric acid evaporates very slowly, especially if the containers are
closed, but if there are losses, they might be replaced by spare acid.

 

[benj]

Unlikely to survive vandalism by the locals.

Even the egyptian pyramids didn't.

Interesting point. Especially since pyramids were mostly nothing but a
huge pile of something as useless as rocks! People still eventually
stripped them down for stones they could use to build a house or
something.

Hence my point of making the storage inaccessible to humans: Mountain
top, Antarctica, deep mine underground, All three, etc. (See millennium
clock...Inside inaccessible mountain.

 

[Bernhard Kuemel]

Interesting point. Especially since pyramids were mostly nothing but a
huge pile of something as useless as rocks! People still eventually
stripped them down for stones they could use to build a house or
something.

Hence my point of making the storage inaccessible to humans: Mountain
top, Antarctica, deep mine underground, All three, etc. (See millennium
clock...Inside inaccessible mountain.

How would I power the LN2 generator there? If it snows on the mountain
top solar energy collectors are probably covered. Antarctica is dark for
several months and receives little sun in the summer. Mines are dark.

Fission/fusion reactors are infeasible to build to work autonomously
1000 years.

I saw a few "mountains" near Alice Springs, 1000+ m high. They are
fairly remote, but not inaccessible, IMO. Algeria has some good spots,
probably, but I think Australia is safer than Algeria.

Bernhard

 

[Arno]

How about radiation hardened parts stored 100 m below ground. Hopefully
not near a uranium vein.

That eleminates the particles, but not the natual atom migration
damage. Sorry, current semiconductors just do not have that
lifetime you need.

We could replace plastic parts with ceramic and use durable metal (alloys).

The problem is that the springs will not make it.

http://de.wikipedia.org/wiki/Langzeitarchivierung#Haltbarkeit_der_Tr.C3.A4germedien
claims up to 700 years for polyethyleneterephtalate microfilm.

"up to" yes, in perfect conditions. But again, the reader will not.

The sun. I'm thinking of using black walls with thermocouples inside
100-500 km north of Alice Springs in the desert.

Thermocouples have the endurance, but how will you keep dust off them?
That will become a problem within weeks.

I forgot about the vanadium batteries (
http://en.wikipedia.org/wiki/Vanadium_redox_battery ). If the membrane
ages, it can be replaced. The battery can even be kept dry until there
is a power loss, then a failing electro magnet will release a weight,
opening a valve and the juices flow to the cell, producing electricity.
The vanadium sulfate solutions have no end of life, IMO. Some evaporated
water could be replaced by water frozen during air liquefication.
Sulfuric acid evaporates very slowly, especially if the containers are
closed, but if there are losses, they might be replaced by spare acid.

If we decide against battery powered service robots on wheels, then a
crane could position robots anywhere in a room. Robots on powered rails
are another option.

Tin? Possible solutions: improved solder. welding. bread boards (new or
old style). cold stored spare parts.

Cold storage runs into the same problems. And no, none of the
other solutions are relly going to cut it, except welding, and
then you will likely need gold or platinum all the way.

Basically you would have to design every elecronisc component
from the ground up to even have a chance. No plastics or
epoxies anywhere. Only ceramics and gold. While doable, it would
be extremely expensive, "billions" are likely not enough.
Just look at what a modern semicondictor FAB costs and you
need one specialized and without the previous generations to build
upon. For example, modern metal layers in chips are aluminium.
That does not cut it for 1000 years. Yet there is no replacement
technology that does. It would have to be developed from scratch.

Arno

 

[benj]

During the same earthquake event, one of my medical office customers had
their file server die because their big UPS kept it running as the
ground was shaking, bouncing the spinning hard disk drive heads against
the platters. (Product idea: Earthquake sensor for UPS). Their backup
tapes were safely stored in a safe deposit box in a local bank vault.
However, that building was red-tagged as unsafe, and entry was blocked
by the same paranoid officials. Burglarizing the bank building was not
a viable option. Access was granted about 10 days after the quake, and
then only through the intervention of high state officials.

Your data is safe, but inaccessible. Now what?

The weirdest disk failure I ever has was from fan moo! Hard disks were in
drawers that had little mini fans on the back. I had already replaced the
original fans with sleeve bearings with ball bearing fans. Nevertheless
after a couple of years one of the fans started to moo. Apparently the
vibration frequency was at some resonance with the heads and banged them
into the platters. I suppose if I paid to have the platters switched to
another drive some tracks might have been saved, but luckily I didn't
need to do that. Hence the backup with multi-copies and keep replacing
dead parts method. Today you can hardly buy a fan that lasts 6 months
before it starts to moo let alone 1000 years! Building fans that last
seems to be a lost technology sort of like how to make a decent mummy.

 

[John Dulak]

Who was it that postulated "No machine may have any moving parts?"

Arno:

That would be Arthur C. Clarke in his 1953 novel "The City and the Stars".

Jjohn

--
\\\||///
------------------o000----(o)(o)----000o----------------
----------------------------()--------------------------
'' Madness takes its toll - Please have exact change. ''

John Dulak - 40.4888ºN,79.899ºW - http://tinyurl.com/3lvoh2n

 

[Martin Riddle]

How would I power the LN2 generator there? If it snows on the mountain
top solar energy collectors are probably covered. Antarctica is dark
for
several months and receives little sun in the summer. Mines are dark.

Fission/fusion reactors are infeasible to build to work autonomously
1000 years.

I saw a few "mountains" near Alice Springs, 1000+ m high. They are
fairly remote, but not inaccessible, IMO. Algeria has some good spots,
probably, but I think Australia is safer than Algeria.

Bernhard

GeoThermal energy production?

Cheers

 

[Bernhard Kuemel]

GeoThermal energy production?

Sounds like there's a chance for volcanism and earthquakes. Gotta think
if that's better or worse than humans. Also I need a cold side to make
electricity from heat. Hot: 300+ m, Cold: 50- m. That's 200 m distance
to bridge with thermo couples or water/heat pipes. Difficult to service
autonomously.

 

[Rod Speed]

Interesting point. Especially since pyramids were mostly nothing but
a huge pile of something as useless as rocks! People still eventually
stripped them down for stones they could use to build a house or
something.

I think in his case they'd just move in and 'live' there.

Hence my point of making the storage inaccessible to humans:

The egyptians tried that too. Didn’t work at all.

Mountain top,

I'm not sure that there is actually even a single mountain top
that no loon has never chosen to visit just because its there.

Antarctica, deep mine underground, All three, etc.

But there isnt any point in doing it unless it
can be accessed after the 1000 years are up.

 

[Rod Speed]

How would I power the LN2 generator there? If it snows on the mountain
top solar energy collectors are probably covered. Antarctica is dark for
several months and receives little sun in the summer. Mines are dark.

Fission/fusion reactors are infeasible to build to work autonomously
1000 years.

I saw a few "mountains" near Alice Springs, 1000+ m high.

There are no hills anything like that high anywhere near there.

They are fairly remote, but not inaccessible, IMO.

No mountains anywhere on the earth are.

Algeria has some good spots, probably,
but I think Australia is safer than Algeria.

Depends on what you mean by safe.

The murder rate is in fact at least as bad as is seen in the
absolute worst of the american urban ghettos in some
of the aboriginal 'communitys' like Port Keats/Wadeye

And most of the land in the northern territory owned
by the govt and leased to the current leaseholders too.

You don’t get to even visit quite a bit of it without
the permission of the 'traditional owners'

 

[Rod Speed]

The pyramids were made from concrete:

Like hell they were. They just use quarried stone.

<http://www.geopolymer.org/category/archaeology/pyramids>
The poured concrete blocks used to make the pyramids were sized as
large as possible to prevent pilfering. Unfortunately for the ancient
builders, they didn't consider the possibilities offered by modern
machinery and blasting explosives.

Difficult to inspect, test, and verify that the data is still intact.
Would pay to store your data without any proof that the data
was actually still intact?

Yes, if its stored well enough like with the engraved nickel sheets.

The more inaccessible the location, the more difficult it is to access.

And there isnt any point in keeping it for 1000 years unless you can
access it again after that time.

Trust, but verify.

But then you have a problem with who is going to do that for 1000 years.

 

[sdrat]

That eleminates the particles, but not the natual atom migration
damage. Sorry, current semiconductors just do not have that
lifetime you need.



The problem is that the springs will not make it.



"up to" yes, in perfect conditions. But again, the reader will not.



Thermocouples have the endurance, but how will you keep dust off them?
That will become a problem within weeks.

No, solar panels work fine there for much longer than that.