Early diamondoid nanosystem pixel (direct path)

From apm
Jump to: navigation, search
Conceptual sketch illustrating the idea of an "early diamondoid nanosystem pixel" or "early replicative pixel" for short. This is an attempt to evade the shortcomings of ultra-compact monolithic-machine molecular assemblers as bootstrapping tools. In order to find a perhaps more viable direct path. See: Proto-assembler (outdated). Evading assemblers as described in the early Engines of Creation (but shunned in the technical analysis Nanosystems), Evading assemblers as presented in KSRM, and evading assemblers as discussed in some designs that supposedly evade them but really just change to "stuck to surface". (See: Discussion of proposed nanofactory designs.) Note that "stuck to a surface" is only a part of the distinction that is insufficient standing alone. – Link to vectorgraphic version.

This page covers considerations regarding an early crystolecule system
that can eventually approach non-compact selfreplicative capability.
This page is not about molecular assemblers.
Specifically not ones in the bootstrapping context: Proto-assembler (outdated).
Also not about molecular assemblers stuck onto a chip like in some old concepts (Chris Phoenix 2003).
See: Discussion of proposed nanofactory designs.

Semi reasonable seeming szenario

Positive formulation:

  • a non-compact system expanding to the size it wants to be in order to eventually reach selfreplicativity
  • an open-exterior nano-system, i.e. there's no vacuum box enclosing an eventually "replicative unit pixel"
    (see vacuum handling below)
  • an non-monolithic open-borders nanosystem (i.e. components from adjacent "replicative unit pixel" can cross over to collaborate)
  • strong separation of concerns in sub-systems e.g. a carrier chassis just to carry subsystems around

Due to non-compactness:

Regarding automation & functional redundancy:
Note that this is by no means an advanced mature factory style nanofactory with yet.
No one-atom-per-station operations and such.
Any functional redundancy from throughput balancing and limited automation is solely to
increase feasibility of buildability and to facilitate faster system scaling.
Early nanoscale quantity sellable products would help in
making this pathway more ecomomically feasible though.
Making some of the above for-scaling-necessary-optimizations easier.

Negative formulation: Unlike an molecular assembler

  • not ultra-compact in volume - not forced in a box of desired size
  • no expanding vacuum hull - and no box enclosing the whole system as part of the nanosystem
  • no monolithic closed-borders design (i.e. adjacent subsystems can cross over)
  • no whole system mobility

Vacuum handling

Only the mechanosynthesis happens in PPV (assembly level 1).
Assembly of crystolecules to assemblies of them (assembly level 2) is done in clean air.
For that a macroscale enclosing cleanroom box (or anything better) suffices.

Vacuum chambers for mechanosynthesis are hard non-expandable and as big as needed.
They can be as big as permitted by the crystolecule assembly mechanism.
Smaller might be preferable for more modularity.

See main page "vacuum handling" for details on expelling the
fully mechanosynthesized and passivated crystlecules out into clean air space.

Vacuum chambers are non-momolithic out of many crystolecules held together by
vdW force or form closing interlock. They are assembled in clean air.

The vacuum just needs a single exponential pump-down using the
lockout mechanism as positive displacement pump.
From then on no pumping is needed because the mechanism has
perfectly zero gas molecule back-flow during crystolecule lockout.
Tunneling is FAPP ignorable.

See: Vacuum lockout

Why not a monolithic vacuum chamber?

A monolithic vacuum chamber crystolecule would be a very big crystolecule.
– Extremely (impossibly?) hard to make wit early SPM systems.
– Also challenging for early nano-robotic mechanosynthesis.

Beside the sheer size of the parts there is the issue hat the stage needs to
fit into the camber yet also mechanosynthesize the chamber.
Leading to wild expanding vacuum camber designs and such.
Generally this is a driver towards more proto assembler like designs
with all the associated problems waiting there.

Also big bespoke monolithic crystolecule designs design may set a bad precedent
when it comes to "design for crystolecule reusability".
See: Recycling and Gem-gum waste crisis

Potentially individually movable subsystems

This is obviously hopelessly incomplete ATM.

Core subsystems

Assembly stages:

  • crystolecule stick-n-place stage
  • mechanosynthesis stage (eventually with gas-tight walls and zero-backflow airlock)

Basic compute – cam follower unit for nonlinear robot control ?? – stepper control stuff? coarse to fine swithcing ??

Crystolecule recomposition management subsystems:

  • part dispenser magazines carried carried on wide attachment chains
    giving a lot of part storage density like in movable libraries but faster access

Very early very simple automation:

  • rail assembly factorylet (crystolecule assembly to standard parts like: rails, struts, chains, …) (think: cranked box acting like a (un)zipper)

Machine elements

These are the sub-systems of the sub-systems.
Typically many off them rigidly connected.
See: Crystolecule based machine elements

Related

External links

Good inspiring sources.
A lot needs to be adapted to the nanoscale physics and context.
Most electrics needs to be replaced by mechanics which is perhaps the most challenging part (and perhaps the most volume consuming part).

Moses2013

Self-replicating blocky-granular gantrybot pick-n-place robots on a square grid of rail-tracks.
See Moses2013

Ambots

An interesting modular system architecture with some relevant aspects:

Note that Ambots has a specialized unit that is just for carrying otherwisely specialized units around that can not move on their own.
Ambots wheels on flat ground would not be feasible for nanoscale at all of course.
Rather than a lot of self contained autonomy in the Ambost case
in the case here (reversible) mechanical coupling to drive chains in a rail network via clutches is needed.
Reciprocative chains for signalling could even be as simple as cuboids in a channel. Not even interlocking.
The trick being using vdW-force and/or push-back spring at the end.

Retrieved from "http://apm.bplaced.net/w/index.php?title=Early_diamondoid_nanosystem_pixel_(direct_path)&oldid=16194"