Difference between revisions of "Self repairing system"

Latest revision as of 14:14, 2 July 2024

The capability of self repair is unlike redundancy not a necessary feature of active gem gum materials and advanced atomically precise products. Self repair can help doing recycling by reducing the rate of waste production [todo: elaborate on that] but also can inadvertently prolong the lifetime of pieces of waste. Further it is obviously prolonging product lifetimes or times of good performance to very long (and hard to predict) time-spans. Added high level repair effectively extendes lifetimes ad infinitum making other limiting factors more relevant (certain natural disasters).

The main kinds of damage include:

• radiation damage
• thermal damage (internal or external)
• mechanical damage
• chemical damage (including weathering)

Different kinds of damage need different treatment.

The simplest form of repair is:

• Feeding the damaged objects back in a microcomponent recomposer device or a complete nanofactory capable of repair.
• There to check which microcomponents are still usable and
• reassemble the original object with microcomponents that where found to be in working order or new ones as replacements for the damaged ones.

Damage that fuses regions of microcomponents together is a lot more difficult to handle (assembly level IV).

Repairing damage live in a product that is under use (e.g. a mocle needs means for internal transport for microcomponents e.g. legged block mobility in a channel network. this further complicates the systems design and may lower maximal performance.

Fundamental types of self repair

In place self-repair

This is defined by that there is no need to disassemble the product with a gem-gum factory to perform self-repair.
For this to be possible there the product needs to have dendritic (tree like) channels

• for broken parts to be removed from in place position and
• for new parts to be supplied to in place re-installation

Since the lowest assembly level with its mechanosynthesis is not necessarily (or not easily) designable to be reversible. The stuff that will be transported through these dendritic channels would most likely mainly be microcomponents or crystolecules. Microcomponent maintenance microbots would take care of local disassembly and reassembly and transport in and out the dendritic resupply system.

The stuff that will be transported through these dendritic channels would less likely be capsules with resource molecules for in place mechanosynthesis. Opting for in place mechanosynthesis would lead to a design quite similar to the (outdated) molecular assembler concept. The very very low rates of turnover necessary for continuous repair of steadily accumulating radiation damage might make the massive slowness of a molecular assembler like design somewhat acceptable though. Maybe.

Live self repair

Live self-repair is in place self-repair plus there is not even the need to turn off active operation of the device (or active metamaterial). There is no need to turn of driving power or block freewheeling motion in e.g. mocles and infinitesimal bearings respectively. Self repair can be performed maximally continuously and incrementally.

In place offline self-repair

While there is no need to disassemble the devise / active metamaterial with that capability to perform self-repair, there is a need to turn off active operation. So the self-repair need to be batched in chunks of downtime which may or may not be perceivable by human senses.

Out of place self-repair / Disassembly reassembly self-repair

This is defined by that there is a need to fully disassemble the product with a gem-gum factory to perform self-repair.
There's a need to disassemble the product-to-repair from the macroscopic product down to as far as the reversibility in the assembly process permits.
Macroscopic automation may or may not be involved. The process seems more fragile and prone to failure than in place assembly for long amounts product maintenance time. over extreme amounts of time.

intermediate forms of self-repair

• External: replacing bigger blocks from the macroscale
• Internal: Replacing microcomponents or crystolecules from the mesoscale.

Countering natural decay

>> Nature always wins and takes back was man has built. <<
For the better or the worse this common saying will slowly loose its truth with the emergence and improvement of artificial self repairing systems.

With weathering including abrasion root growth and UV radiation there are chemical mechanical and radiative damage sources. ...

Diamond is rather resilient to bases and acids. Abrasive damage through silicate dust carried by the wind is more likely to do damage. ...

Thermal damage

Thermal overexposure of macroscopic volumes (singed spots) need those volumes to be disposed and replaced. (microcomponent damage crop-out)

If there are means for microcomponent disassembly (reversible locking mechanisms are used) each microcomponent can be tested for their functions which they expose. Furthermore microcomponents may provide a functionality to test some internal functionalities. Still not everything can be tested some displaced atoms or structures may be undetectable and cause failure at a delayed point in time. Even the testing functionality itself might be broken. Very speculatively and not seriously considered here one might try to use TEM (transmission electro microscopy or gentler future de-broglie-matter-wave microscopy) to check all the atoms positions but that itself might induce damage and/or take too long.

In essence one never can know for sure whether a microcomponent still has all atoms in place (is AP) or not. Thus the singed to non singed border layer between still usable and damaged microcomponents is somewhat fuzzy.

After self repair the outer part of the fuzzy damage border layer shell remains in the original product or got to be reused elsewhere. When repairing external thermal damage one has to draw a line (shell) between the microcomponents that ought to remain and the ones that are to be disposed (burnt). Reusing microcomponents that where too close to the thermal damage can cause some kind of "invisible damage poisoning" so its generally better to keep your space to the highly damaged area and generously cut away and dispose of microcomponents instead of reusing all the ones that still fulfill their external function tests at the time of scavenging / repair.

To decide whether to reuse a microcomponent or not the integration of a thermal seal might be useful. (recording thermal history) When scavenging microcomponents from the vicinity of a damaged area an internal pristinity switch could be flipped or a usage counter incremented.

Getting out the highly damaged macroscopically fused block is another issue - possibly requiring macroscale robotics.

Mechanical damage

Assumptions:

• Something inside an advanced gemstone based nanosystem breaks (purely internal mechanical damage). This might be due to a software bug, a not properly handled overload, or another reason.
• A clean break with perfect or near perfect cleavage (nothing like a massive radiation hit mixing lots of atoms up very intensely).
• The damage did is purely internal and did not break the seal into the PPV machine phase environment.

Check then repair and wastefulness

Let's assume a structural strut or some part of a housing structure that is made out of many reusable subunits breaks somewhere. From everyday macroscale experience one is used to seeing the damage and simply replacing that single broken part. But in nanosystems "seeing" requires exotic non-optical methods (TEM or even helium matter wave microscopy) and likely will remain exclusive for rare super in depth investigations.

Instead of "looking" one rather needs to touch test the structure all over with specialized devices to check whether everything is in exactly the state one assumes it is. (Preferably the same tools that are used for assembly to keep complexity in check). Weak spring locks that jumped into an faulty state (by the odd strong thermal fluctuation) could be seen as a mild form of mechanical damage. for these damage states its known how to check for. They are easier to detect and correct. Just shift/turn/.. back into correct state.

Since this is time consuming in both touch test system design and execution (actual touch all over test). Its more than likely that this will often be skipped and only be pursued in retrospect when it really becomes necessary.

Depending on the location of occurrence a crystolecule breakage damage might go unnoticed for quite a while or even permanently. (which may lead to the assumption that some certain low performance is just normal albeit in reality its just due to some really stupid bug.) What is very likely is that if the error is detected it is detected as consequence down the road (maybe when many identical things broke in identical situations and the problem became really serious and pressing).

It might be relatively easy to detect that there is some sort of damage in a specific type of microcomponent and then filter them out and send them of to "thermal recycling" (euphemism for burning).

But 99.9% of the crystolecules inside these microcomponents in question might be perfectly fine though. So disposing of them is pretty wasteful.

Nonetheless things like this will most definitely happen. Software development has shown that in highly automated systems (where some resources are almost too cheap to meter) there can even be some merit in accepting to have extremely stupid naive and wasteful "bugs" manage hide away for long periods of time. Good example compiling high level languages into machine specific assembly code.

In physical systems this kind of wastefullness can have a bit more serious consequences though. Disposed of physical material does not magically vanish like data in ephemeral RAM memory. In a worst case scenario a physical analog to a digital "space leak" (uncollected data garbage => uncollected microcomponent garbage) this might even take the magnitude of a microcomponent pipeline breach.

Internal spill

Assuming storage container breaks that is holding some crystolecules (e.g. bearings) After such a break the crystolecules are not anymore fully enclosed (no more restrained by shape-locking). At moderate temperatures (including room temperature) the crystolecules are very unlikely to spill due to the fact that even very minor contact area to the housing provides significant VdW binding energy preventing that.

In case that there is:

• low contact area
• very high temperature
• the one odd unlucky thermal fluctuation spike

a crystolecule may occasionally really get spilled (into the still sealed environment).

One might worry that it may act as wrench in the gears. The places where a spilled crystolecule is most likely to end up are likely corners where all three degrees of freedome of translational movement are suppressed by VdW sticking to the walls. Depending on the system in question voids might be

• big with lots or corners and few moving parts inside or
• small with few corners and lots of moving parts inside or

In the latter case its obviously more likely that a spilled crystolecule meets moving parts. If it comes into moving parts in some cases it might not pose a problem due to the extreme slipperyness of crystolecules (e.g. probably so the case when coming between small gears). Other cases are serious (e.g. sticking in the corner of pistons preventing full displacement)

Notes

(wiki-TODO: Discuss distinction between "in situ self repair" and "offline self repair" and their relation to recycling and spill)

Related

• Consistent design for external limiting factors
• Microcomponent maintenance microbot
• Microcomponent tagging
• Debugging
• Recycling and Spill
• Redundancy
• Radiation damage

(wiki-TODO: Add auto detection that product is abandoned - no usage like accelerations - timeout - dead man's button)