Archive for the 'inventions/schemes' Category

RFO: One-way book-to-PDF scanning service?

May 12, 2009

(RFO == Request For Opinions)

I’ve been doing some traveling lately, to the east coast and to Chicago, and found that I still like it.  I’m not 20 anymore, but I can still deal with some hassle and bullshit from time to time and appreciate the surprises and adventure of going to new places and learning stuff.

(And things like WiFi, iPhones and make it so much easier, it ain’t funny.)

So I think about how I could do that as much as possible, or even “move to nowhere” for some years, as I’ve been fantasizing for some time.

And here’s the catch: Books! What about my books!

I can pare my life down to five each of a certain shirt, underwear, pants and shorts.  A certain blue blazer and pair of shoes.  A MacBook and certain fiddly accessories.  That’ll pack my world down to a (waterproof, airline-checkable, shotgun-resistant, rollable, lockable and enviably macho-looking) Pelican 1620 hard case. Great!

But the books!  The books!  Dangit!  I want access to my books!

So how about this.  It lets me “dispose” of my books physically without saying goodbye entirely:

It’s a super-simple streamlined business that receives books (and magazines), slices off the bindings, sheet-feeds them into a pack of PDF’s, and emails them back.  (‘And implements the ~100 non-obvious details that make it real and genuinely useful.) They’re legible on one’s laptop or, even better, on one of those easier-on-the-eyes electronic-ink Kindle thingies.

The books are always destroyed!  Bye-bye!  That makes the service cheap and fast.  Five cents a page?  Four?  Three?  Two?

Services sorta like this exist, but their websites are complicated and crappy (== government work).  No one’s really ironed this out to something that’s as slam-wham-bam as it should be, given the assumption (that no one’s ever made) that we’re only doing one-way scanning, and never two way.

My question: Am I the only freak who can see himself using something like this?

Triumph of the obvious: Bollards: A victory for ParaTow!

March 3, 2009

So.  I’m super-duper into this ParaTow wind-based ship-propulsion scheme, right?  Right.

And a remaining mystery, to me anyway, was the question of how the hell to hook it to a ship in a safe and inexpensive way.  I was assuming that some kind of harness contraption would have to be bolted/welded/strapped to the boat to allow the ParaTow to attach to the ship in one place.

And that’s probably true… if I wanted to hook to the ship in one place.


Take a look at this.  There are these thingies on ships called bollards.  They’re called bollards whether they’re part of a ship or part of a dock.  They’re the hard points where you tie on ropes or chains.

Bollard (I think on a ship, but I can't tell for sure)

Bollard (I think on a ship, but I can't tell for sure)

Bollards on a dock.

Bollards on a dock.

Okay.  So.  I finally decided to ask just how much load these things were built to take, and then compare that to how much tension a ParaTow will be exherting on a ship to pull it around at full speed.

I couldn’t figure that out, actually, but I did figure out that the “bollard pull” of normal harbor tugboats (how hard they can yank on a ship’s bollard through a long rope) is around 50 tons.  Okay.

Further, the bollard pull of an oceangoing tugboat (basically, a tow truck for ships that break down at sea) is around 100 tons.  Okay-okay.

So great.  We know that ships’ bollards, and therefore the structure underneath supporting them, are built to take somewhere between 50 tons and 100 tons of force safely.  Okay.

So how many tons must a ParaTow exhert in order to pull the ship at full speed?  If power = force x speed, then force = power/speed, and the million-dollar linux utility “units” will spell it out for us.  For reference, a common Panamax cargo ship goes about 20 knots and has an engine no bigger than 50,000 horsepower, so:

You have: 50000 hp / 20 knot You want: tonf * 407.33261 / 0.0024549962

Aha, so about 400 tons of thrust.

Well then, if a ship’s bollard is easily good for 50 tons of oomph, then we’d (theoretically) only need to hook to eight of them (four on the left side plus four on the right side) in order to safely yank this thing along at full speed.

The good news is that according to the pictures I’ve found in Ships Monthly magazine, they all have at least ten bollards up front.

So what do you know, there it is!  I’m imagining some kind of super-tugboat platform with not just one winch and tow line, but ten or so.  Their relative lengths are tuned to the geometry of the given customer vessel, and it goes through some procedure to let the ship haul up each of them and loop it over the bollard in question.

So wow.  The real world is never as simple as the sixth-grade math in a casual blog post, but the basic message is still loud and clear: When working together, the bow bollards on a ship are together rated for the same range of cumulative line tension as a ParaTow needs to route into the ship’s structure.

So done intelligently, it’s looking possible to demonstrate the ParaTow concept at full scale without having to add new structure to the ship (what I think most of you have suspected all along, but I was trying to make complicated).  Wow!  Let’s hear it for logic!

Container Blimp

January 7, 2009

Instead of trying to justify the building and operation of very large blimps by exploiting their unique VTOL (vertical take off and landing) capabilities, here’s a different approach.  Let’s look at finding a niche for blimps in an established, commoditized and analyzable industry: container shipping.

There’s a “hole” in one’s options to ship a container across the ocean.  It’s either on a super-slow and super-efficient ship or on a super-fast and super-gas-guzzling jet plane.  The fuel-burn ratio is like 100:1.

There’s no middle ground, and economic dogma states that The Market always wants to be segmented.

o Mac Mini, iMac or Mac Pro?  15″, 17″, 19″, or 21″?
o Chevy, Pontiac, Buick, Oldsmobile or Cadillac?
o Old Navy, Gap or Banana Republic?

Thesis: Some containers are late, and cause losses due to downtime at their intended destination.  The only choice in these situtations is to eat the downtime losses or repack everything into a cargo airplane and pay through the nose to get it there on time.  Either way, it’s expensive, and that expense is bourne by the economy in various ways.

A Container Blimp, then, is that middle ground.  It has a speed and fuel efficiency somewhere between a ship and an airplane, truck-ish or train-ish.  Surely my math is simplistic when I try to locate the correct size, but it’s definitely there somewhere.  Blimps double their fuel/ton-mile efficiency with every doubling of scale.

Besides conventional container traffic, such a contraption would also allow “UPS Ground to China.”  Just three days across the Pacific Ocean (or two days for the Atlantic), for instance, would let “about a week” thinking still work when shipping packages to foreign countries.

And an added trick is the idea of not landing the blimp at some airport and having stuff loaded into it, but rather to hoist it up and set it down all at once in the form of a special-purpose barge (exchanging it with water mass each time).  That way, interfacing with the existing infrastructure (cranes, trucks, railroads) is entirely conventional.  Also, it means the blimp is never flying over land, which will surely help regulatorially.

Parafoil Dynamic-Ballast Blimp; take 2

January 3, 2009

(This is a next-level follow-up to a previous posting.)

Over the break I found my sliderule and got more serious about estimating this thing’s size and performance.

I dropped the idea of using methane (CH4) as a lifting gas and instead went with just hydrogen (H2).  Hydrogen can be made from diesel fuel (vital for military markets… sigh), lifts more per cubic meter and has a lower fuel energy density (about 75% less), which is exactly what we want in this case.

See, the backbone of this whole scheme is that instead of having to vent lifting gas upon putting down something heavy, it allows you to productively recover its fuel value over the return trip (if it’s a certain distance or longer).

It uses deployable hanging parafoils to pull down on the blimp after putting down something heavy.  Its propulsion engines switch from a lift/weight equilibrium of hydrogen and diesel to hydrogen only.  Ergo, the blimp’s static lift is contantly decreasing until the downward-pulling parafoils are no longer needed.  Once enough hydrogen has been burned the parafoils are retracted back up against the envelope (and out of the airstream), and the engines switch back to hydrogen and diesel for the rest of the return trip.

Okay.  So far so good.

What I learned is that the parafoils-deployed hydrogen-burning first part of the return trip is still 800-1000 miles long.  To set down an interesting-size payload (a 40-ton container or a 80-ton M1 tank), that’s still a lot of hydrogen gas to burn up in the engines, and it takes a while.

And then I found that most interesting off-road shipping routes (Kuwait<->Badhad, ‘Stans<->Kandahar, Yellowknife<->the diamond mines in northern Canada) are less than 1000 miles each way.

That’s a bummer, because the blimps need a return trip 800-1000 miles long to profitably burn up all that hydrogen.

Cheat #1: Just vent the rest.  Doing so torpedoes the scheme’s whole raison d’etre, though.  If you’re usually going to vent ~half of it then why not vent all of it and skip out on the deployable parafoils complication entirely?

Cheat #2: Gain weight along the way back by scooping up water from a lake or ocean.

The bummer about water-scooping is that it’s dangerous.  The snorkel could clobber a person, boat, fishing net, sandbar or iceberg.  Ergo, it’s much more expensive to do as a robotic UAV (unmanned aerial vehicle).  A UAV can go from point A to point B with no problem as long as both points are up in the air, but to buzz the surface takes skill.

As long as it works, though, water-scooping is always a possibility when delivering cargo straight from a ship or barge.  There are probably a number of instances where this ability to operate “port-less” would be a big winner.

OTOH, if the route is entirely over land, with no bodies of water handy, one would instead have to dig long water-filled trenches somewhere handy for the blimps to scoop from.

So yeah. I’m a little bummed that the water-scooping complication is turning out to be necessary, but I don’t think it’s a show-stopper yet.  More later.

Hanging-Parafoil Dynamic-Ballast Blimp!

December 21, 2008

(This posting was followed up in a more analytical way here.)

Here’s a blimp that can put down a payload in a remote place.

This is actually a big deal.  A conventional blimp can carry a heavy load much more fuel-efficiently than an airplane or helicopter, sure, but can’t put it down!

When the weight of the payload is lost, a regular blimp suddenly has too much lift and can’t land!  Dangit!

The only two (believable) solutions to this problem so far even suggested, as far as I know, are to either 1) Just release some expensive lifting gas into the atmosphere, or 2) Trade the payload for something equally heavy like water pumped up from waiting tanker trucks.

Both solutions #1 and #2 above stink.

#1, venting lifting gas, stinks because lifting gas (helium or hydrogen) isn’t cheap.  If I want to put down a nine-pound gallon of diesel fuel at a diamond mine in northern Canada, but also have to vent four or five dollars’ worth of lifting gas in the process, then that diesel just got really expensive!

#2, trading the payload for water pumped up from below, sucks because not only must there be water trucks handy, but they must be able to get to the drop site. If a truck can get there then what’s the blimp for?

So!  Here’s my solution #3:

Right before dropping its payload, the blimp deploys these hanging parafoils (like the wings of paragliders, but hanging upside-down) and gooses the engines to start moving through the air.  The hanging parafoils generate aerodynamic down-force (like aerodynamic lift but in the downward direction) to compensate for the payload’s lost weight!

The key complication to this scheme, then, is that in order to be able to stop again — which has to happen eventually, namely back at the logistics base where the mission started — the blimp needs to lose lift and/or gain weight, so that the parafols won’t be needed anymore and can be winched back up against the blimp’s envelope.

Lift can be lost in a productive way by having all or some of the lifting gas be CH4, aka methane, aka natural gas, and just burning it up as fuel (‘cuz ‘gotta burn something, and CH4, if it’s available, is always cheaper than diesel fuel anyway).  Further, water vapor in the engines’ exhaust can be condensed to liquid water and retained in tanks, thus doubling the lift-losing/weight-gaining effect.

(The same scheme works if the lifting gas being burned up as fuel is just hydrogen.  You just need the extra machinery at the base to make that hydrogen gas in the first place plus engines that can burn it safely, a perfectly doable but not-so-off-the-shelf sort of thing.  What’s nice about hydrogen, though, is that not only does it lift twice as much as CH4 per volume, but gives about a quarter as much fuel energy per volume too.  Ergo, burning hydrogen gas instead of CH4 lets us lose-lift/gain-weight about 4X as quickly!)

Another way to gain weight is to fly over a body of water at any point along the way back and scoop up water through a special boom.  Here’s a video demo of a helicopter doing this.

And there we have it: A blimp (with some physical and procedural complexities stapled on) that actually can put something down in a remote place in a controlled, cost-efficient and non-catastrophic way!  This could be a big deal in the oil/gas exploration, military logistics and/or emergency humanitarian aid businesses (and/or the other fields that you can help me brainstorm here).

A final interesting tidbit is that since it’s so advantageous to be able to scoop up water on the way back to base, it’s an interesting idea to have that base be a ship.  That’s pretty interesting, becauase now you’re doing heavy deliveries in-country directly from a ship, without having to deal with the roadblocks, washouts, mudslides, theft, spoilage or bribe-seeking customs goons on land!  Awesome!

$6M of venture funding for WHAT?

December 12, 2008

I swear to God, there’s just so much (well-marketed) bullshit out there that sometimes I just can’t take it.

“I know! Let’s quintuple the part count and material cost per watt in exchange for a 50% better efficiency!”

“Awesome! Here’s some money! We’ll pass it off to someone richer and dumber later!”


November 17, 2008

All right!  Another biggie!

The FlyMill has been a big big part of my life for years.  Here’s a 30-second video about it that I made for my Google Project 10^100 application:

The result of many years of crazy-person-esque obsession and anguish, the FlyMill was (and perhaps still is) the very best I could do at coming up with a quite-scalable scheme that would need as little physical strength per watt as conceivably possible, so as to deliver super-cheap renewable electricity in the ballpark of 2 cents/kw-hr (if all of its many engineering, mass-production and logistical challenges were solved).

(It also has some very serious problems, even as an idea, which I’ll save for last.)

So it’s basically a set of “electric airplanes,” with direct-drive windmills for propellers, booking around in a circle.

The planes are made of metal or plastic, and not cloth like some other kite-power schemes I’ve seen.  I just can’t believe in anything that’s not clothless, because no cloth lasts very long in the open air 24-7 (without maintenance).

Super-genius autopilot flight control lets the planes bias their tether up above horizontal, and thus not crash into the ground.

When there’s no wind at all then the tether points straight up and the airplanes consume electricity to stay aloft, because the idea of landing and then taking off is just too horrifying for me to think about.  They consume some fraction as much electricity to stay aloft in dead air as they produce when there’s a decent wind.

(A windmill, when driven instead of driving something, makes a crappily-inefficient but still somewhat-functional propeller.  That’s how the electric airplanes can theoretically propel themselves at low speed.)

Since it’s tethered, it could be deployed out to sea.  Chains get linearly more expensive with length=depth, while with windmill towers it’s something like the third power of length=depth.

The abiilty to cheaply work in deep water is a big big deal.  There’s a lot of wind and a lot of real estate out there, and that real estate is far from most authorities with the power to sue. Out of city waters, out of county waters, out of provincial waters, etc.  This is important!  Terrestrial wind power is basically illegal in France, for instance, because there are so many grounds (environmental, eyesore etc.) upon which someone can sue to keep a wind farm from being built.

Another key idea: The last 20% of a regular windmill blade’s length is where 50% of the power is, but is also the cheapest 20% of the blade’s length.  A regular windmill blade must get stronger and stronger the closer it gets to the hub, which is where the material=weight=expense is.  The FlyMill has blade tips only!

So on a per-watt basis, this tips-only property is why I believe that a FlyMill needs less material/physical strength per watt than a conventional windmill.  If everything else about it can be made to work and Detroit-scale mass production (hopefully not involving carbon fiber or even aluminum if possible) can make the planes, then this lower material-per-watt ratio would be the ultimate key to the FlyMill’s low cost.

The FlyMill is also gearless.  The “propellers” direct-drive the motor-generators.  This is important because the gearboxes on regular windmills are frightfully complicated, expensive, and still keep breaking down!  The FlyMill has no gears!  Yes!

So.  It can be installed over more real estate, in better winds, while using much less material strength than regular windmils.  So, why aren’t they all over the place by now?

Well, dammit, because of some very ugly apparent showstoppers.  Showstoppers I just haven’t found a way around, even in the imagination:

Showstopper 1: The autopilot control algorithm of the planes will be super complicated.  Just one crash into the ground and it’s all over.

Showstopper 2: The electric airplanes have many actuators.  They have rudders, ailerons, tail flaps, etc.  So how oh how could they keep on actuating, day in day out, for 10+ years without being serviced?  And if not, how could they be serviced? I don’t know!

Showstopper 3: The power electronics needed to speed-control the many “propellers” on the planes and combine their outputs into a single high-voltage cable to the ground would not be free.  Furthermore, the power cable coming down along the tether will be prohibitively heavy unless the electricity is upped to many thousands of volts.  That’s not free either.

(Am I making any sense here?)

Armageddon Calculation, revised

November 11, 2008

So dig.  There are some very important chemicals — some of them viable piston engine fuels as well — that are conventionally made out of oil & gas that can instead be made with air, water and renewable electricity.  At a certain price-point for the electricity, their “synthetic” renewable versions become price-competitive with the conventional fossil stuff.

For instance, ammonia (NH3) is a big-money industrial chemical.  It’s used to make artificial fertilizer, for one, and is also a viable piston engine fuel that’s storeable at room temperature under propane tank pressures.

The N in ammonia is distilled from the air, and the H is typically extracted from natural gas.  However, both of these elements can be obtained from air and water with cheap-enough electricity.

At a rough wholesale price of $200/ton, and a heating value of 18.6 mega-joules/kg, that means that the heating value of ammonia is going for 4 cents per kilowatt-hour.

Ergo, simplistically speaking, if someone could generate renewable electricity for 2 cents/kw-hr and use it to drive a 50%-efficient process for making ammonia from air and water, then the resulting product would be price-competitive with the stuff made from (air and) natural gas.  This assumes that the factory is free, which of course isn’t true, but I’ve read over and over that when it comes to making ammonia, the dominant cost is the gas feedstock, not the capital payments on the factory itself.

So.  2 cents/kw-hr could lead to a monster industry that displaces demand for natural gas and liquid motor fuel.  These are some mighty-big revenue streams we’re tampering with here.

Furthermore, let’s talk about methanol (CH3OH), an even-better motor fuel because it stores at room temperature and atmospheric pressure.  Methanol can theoretically be made from hydrogen and CO2, but no one’s yet had reason to do this on an industrial scale.

Well then, hydrogen can be obtained by electrolyzing water, and CO2 can be extracted from the atmosphere (or captured from the exhaust of a gas or coal power plant).  I posted a little earlier about a new process that allegedly extracts CO2 from the air at an energy cost of 1000 kw-hr/ton.

Ergo, one can imagine a process that sucks in water and air, extracts the hydrogen and CO2 from them respectively, reacts them somehow via the “reverse water gas shift reaction,” and spits out methanol.  From what I’ve been able to calculate, if the 1000 kw-hr per ton of air-extracted CO2 figure is for real, and if the rest of the process were 100% efficient, then the electricity feeding this process would have to cost 3 cents/kw-hr in order for the resulting methanol to be price-competitive with diesel fuel (the heating value of which goes for 4 cents/kw-hr as well, just like ammonia, which is interesting).

So.  The Point I’m trying to make here is that agriculture (via ammonia) and piston engines (via ammonia or methanol) can be economically driven by renewable electricity if that electricity is cheap enough.  No batteries, no fuel cells.  If these processes can be made 50% efficient, then “renewable ammonia” could be made at 2 cents/kw-hr, and “renewable methanol” at 1.5 cents/kw-hr.   (Roughly, and assuming the factories are free.)

Those are low numbers, but at least not comically low.  If anyone can actually crack them in a scalable way then it looks like they’d be creating a multi-trillion-dollar industry.  Not only would it displace where we get our electricity from (that happens at a much higher price point, around 5 cents/kw-hr), but our fertilizer and motor fuel as well!

That’s a big damned business!  Way bigger than Google.

BTW, for your reference, 1 cent/kw-hr = 8.7 cents/watt-year, or ~50 cents/watt over five years or ~$1/watt over ten years.  That gives a rough idea of what the installed wattage will have to cost in order to make such low electricity prices possible.


UPDATE: We’ve all heard about how gasoline cars can be converted to run on methanol, so that’s an obviously-done thing.  But what about ammonia?  Aha!  As further proof that ammonia is a perfectly-workable motor fuel as well, watch this video about a man who converted his ’81 Impala to run on it and is still driving it that way!


UPDATE #2 (Nov 17 2008): Thanks to some very good contributions from you kind and handsome commenters, I’ve learned about a still-experimental but very promising processed called SSAS, or “Solid State Ammonia Synthesis.”  It’s basically an ammonia-powered fuel cell driven in reverse.  Nitrogen + water + electricity in –> Ammonia and Oxygen out.

This is a big deal for many reasons:

R1: It uses about 40% less electricity than an electrically-driven Haber-Bosch process (look up Haber-Bosch in Wikipedia).  Quoted/estimated numbers are about 60% efficient when compared to the (unattainable) ideal, which is great because I was hand-waving/hoping for 50% efficient above.  This savings mostly has to do with the fact that SSAS doesn’t make hydrogen gas as an intermediate step.  That’s good, because otherwise, in Haber-Bosch, the reaction of Hydrogen and Nitrogen to make Ammonia is a exothermic one, thus blowing some of the energy it took to make that pure Hydrogen gas in the first place.

From what some experts have estimated, 2 cent/kilowatt-hour electricity feeding SSAS would produce ammonia at about $220/ton = $1.75/equivalent gallon as a motor fuel, which would rock the house!

R2: It has the potential to cut way down on the up-front cost of an electricty-to-ammonia factory because it all happens at solid state and low pressure.  There’s much less pumping, expanding, piping, boiling, condensing, separating and heat-exchanging going on.  The reaction chamber of a Haber-Bosch reactor works at 200-300 atmospheres (3000-4500 psi), which doesn’t come cheap!

R3: It flattens most economies of scale.  A SSAS ammonia factory of output “10 units” costs basically 10 times that of a factory of output “1 unit.”  It’s basically linear.  The efficiency is pretty much constant across scale as well.  This means that small electricity-sinking ammonia-making factories would be as economic as big ones.  (SSAS plants of most sizes would surely be made of parallel-running shipping-container-sized units of capacity, which is exciting for the mass-production benefits of affordability and scalability.)

R4: They can work intermittently.  A Haber-Bosch reactor is apparently a necessarily always-on kind of thing, like an iron smelter in a steel mill.  Turning a Haber-Bosch reactor (safety) on and off is a days-long production.  SSAS, on the other hand, can be throttled in a second.  This is perfect for sinking variable and intermittent power, like from a wind farm or the otherwise-unused nighttime baseload output from a dam or other power source that feeds a varying load but is barely throttleable itself.

So how about that!  What’s neat about this is that it points to the business model of setting up windmills in remote un-grid-connectable but hella-windy places (like Patagonia, Alaska, Tierra Del Fuego, etc.) and using that resultingly cheap but variable electricity to make ammonia.  A truck or ship lumbers up occasionally to empty the stationary tank and haul it away to market.  Once SSAS is fully real the above is a solid play.

(Also, it looks good for sinking cheap electricity generated far out to sea, which I’ll get to later 😉 )

“PCC” (Parallel Compound Cycle) aircraft engine

October 21, 2008

This is me trying to find a way around (some of) the complexity that doomed the Napier Nomad of 1947.  The Nomad, now a museum piece, is still the most fuel-efficient aircraft engine ever built (in terms of simple mechanical output per fuel input), but it didn’t work very well and came out right before big dumb lightweight turbojets and cheap post-war oil:

(below is a video, not a picture)

Like ye ole Nomad, it’s basically a gas turbine with a very-compressed diesel engine for a combustor.  Since a diesel can handle significantly higher instantaneous maximum temperatures than a turbine can, better fuel efficiency results.

Plus, the gas turbine also acts as a bitchin’-powerful turbocharger for the diesel and keeps its size=weight down when compared to a regular diesel running by itself.

And finally, it’s apparently a better deal to let both the diesel and turbine produce power, as opposed to a diesel with a simple turbocharger that only drives itself.

But, instead of using scary gears to join the gas turbine and diesel together onto a single output shaft, the Parallel Compound Cycle engine cheats.  No gears.  Each side, the diesel and turbine, drives its own propulsion fan.

So the diesel and turbine sides are connected only by tubes (compressed air and exhaust to/from the diesel), not gearing.  They’re both propelling the same airplane, so their power outputs are combined that way.

Further, since there are two “actuator disks” per engine (one for the diesel’s fan, and another for the turbine’s) instead of one, propulsive efficiency (thrust per horsepower) is upped a little bit to boot.

This was probably written about and shot down decades ago but dangit, I just haven’t found anything about it.  Just in case this is a new idea, I offer it up here.  Thank you.

The 20 Year Computer

October 10, 2008

This one hasn’t let me go for months.  The idea is to zealously over-engineer the few parts of a computer where almost all failures happen: the power suppy, disks and thermal management components.

Not “hot-swapping” bad parts, but rather just building in the spares it’ll need down the line and forgetting about it.

The point is being able to deploy a bit of software to a site and not be constantly biting your nails about it failing because the computer broke.   There are $2k computers out there controlling $2M machines with a some-hundred a month service contract, which just doesn’t make a lot of sense.  I’m interested in whether there’s a niche out there for $3k computers re-packaged into $15k bunkers that do the exact same thing, but for much longer, unattended.

There’s “cost-effective” and then there’s “trust-effective.”

(A point I forgot to make is that this thing has a slow-clock low-power supervisor computer, with its own mil-spec Vicor power supply and little UPS, that handles the temperature sensing, fan control and power supply switching.  That’s custom, and is invisible to the big “other” computer.)