Bootstrapping methods for productive nanosystems
A nanofactory needs the parts of a nanofactory to make another identical or improved nanofactory.
There's a chicken egg problem here.
The solution to this dilemma is not ultra compact self replication like in the outdated concept of molecular assemblers.
The solution to this dilemma is a gradual improvement on the parts of currently existing technology that are most critically necessary and most relevant for getting to the target technology as soon as possible. So the task is identifying these technologies and improving on them in a focused way.
Doing so eventually may lead to a self emerging highly distributed self-replicative capabilities of the system as a whole. Similar as human technology is self replicative a a whole. Just that this will be all on a single chip.
Contents
What do we want to do - What do we need to do
To produce macroscopic amounts of atomically precise structures that
- do not only have nanoscale feature sized (like simple molecules or protein medicines - we can do that today) but rather eventually
- do have (atomically precise) features at all sizes scales all the way up to the macroscale
… there needs to be a combination of:
- an introduction and eventually exponential growth towards massively parallel means of manufacturing.
- an establishment of full location adressability over increasingly large size scales. Growing repetitive crystals is easy.
Beside that stiffness of used materials need to be continuously improved to:
- get positional atomic precision not only topological atomic precision
- get to high performance materials and swift production speeds
There are at least three independent orthogonal axes where technological capability can be judged by and scaled along. These are:
- Convergent selfassembly levels
(like experimentally demonstrated in SDN: bricks to blocks then blocks to multiblocks) - Material stiffness
(SDN, protein, stiffer stuff) - Degree of introduction of positional assembly aspects
Bottom Up
Bottom up thermally driven self assembly
Thermally driven assembly excels at massive parallelism but
today's artificial thermally driven assembly is very limited in
the second critical requirement that as already mentioned is: total adressability over large size scales.
One of the prime first objectives is thus to scale this addressability for thermally driven assembly.
In this regard there has been some progress in convergent self assembly of structural DNA nanotechnology.
Stiffness of these SDN structures is very low though and there is no positional atomic precision.
There is only topological atomic precision.
Further bottom up approaches
Beyond only thermally driven selfassembly alone. That is with mixed assembly methods.
Foldamer 3D printers
This includes the scaling approach of expanding the kinematic loop
which leads to the concept of foldamer printers.
foldamer printers would requires already sufficiently scaled up thermally driven assembly such that
- such foldamer printers can be self-assembled
- parts for these printers to semi positionally assemble can be pre-self-assembled
foldamer printers would enable:
- the printer making parts from better materials for an eventual next gen printer
- the printer making parts for eventual larger positional assembly structures
- In the foldamer printer concept there is both thermally driven assembly and positional assembly involved.
- Eventual introduction of [[diffusionless selfassembly] for
The nanoscale foldamer based printers being fully self assembled would make production of such devices scaleable. Thereby giving some positional assembly skills at scale. But the scalability here is without large scale addressability.
Foldamer pick and place robots
foldamer pick and place robotss would requires already sufficiently scaled up thermally driven assembly such that
- such devices can be self-assembled
- the parts for these devices to pick and place can be pre-self-assembled
An idea here is to go for less compact replication on the second assembly level.
rather than the ultra compact self replication on the first assembly level that
is present in the outdated concept of molecular assemblers.
See: RepRec pick and place robots
If parts eventually become big enough in the assembly hierarchy of mixed convergent assembley then direct manipulation with top-down manufactured MEMS devices might become possible.
Printer or pick and place robot - which first
This is hard to say. These two seem to lie on two somewhat orthogonal axes on the technological capability landscape. Both require a focus on scaling technological capability in thermally driven self assembly.
- Foldamer 3D printers - put a focus on material stiffness improvement and material diversification
- Foldamer pick and place robots - increases the focus on scaling technological cabability in thermally driven self assembly
List of non self replicating bootstrapping tools
- bottom up: fully parallel (and hierarchical) self assembly of atomically precise chemically pre-produced building blocks
- top down: conventional photolithographic methods (MEMs)
- exponential assembly – somewhat of an oddity – the glue between bottom up and top down? – time will tell
List of non self replicating scenarios
- non compact self replication – as self emerging highly distributed self-replicative capabilities
- medium compact self replication on the second assembly level
- ultra compact self replication via molecular assemblers (outdated since hellishly difficult and very inefficient)
See main page: Self replication
Related
- Up: Where to start targeted development
- Bootstrapping atomic precision
- Pathways to advanced APM systems –Incremental path and Direct path
- Bridging the gaps
- Thermally driven assembly
- Exponential assembly
- Molecular assembler – not a good bootstrapping method
