3 ms·
I'm so jelly! ("jealous") I've been (slowly and painfully) working my ways towards constructing a self-reproducing robo-boat swarm for a few years now, ever si
by carapace 9y ago
I'm so jelly! ("jealous")
I've been (slowly and painfully) working my ways towards constructing a self-reproducing robo-boat swarm for a few years now, ever since learning about the Great Pacific Garbage Patch.
My idea is to set up a simple artificial ecosystem of interacting bots that collect and recycle oceanic trash to create more of themselves. This permits exponential growth, making it realistic to attempt to tackle such a large-scale problem from a relatively tiny "nucleus".
I want to use Molten salt oxidation [1] to convert trash to "synthesis gas" [2] then feed that through a water column to convert CO+H2O to CO2+H2. Recover the hydrogen for power (more power, the MSO reaction is exothermic) and feed the carbon dioxide to algae tanks. The algae can be processed to create both plastic membranes and a kind of ocean-water-proof glue. Rather than build boats, you make spheroids out of the plastic and glue them together to make foam. (I call it "spittlebug architecture" [3].) Foam and tendons allow you to make e.g. artificial hydras [4] which can be attached at their bases to long conduits (also make of glue-foam) to collect and sort trash and carry it (via peristalsis) to the MSO unit. The long hydra-lined conduits can be arranged in a (huge) spiral which can then be driven to rotate by positioning of sails and rudders embedded in the conduit. The MSO unit would be in the center.
The idea, at every stage, is to do and use the minimum of action and materials and energy throughout the system to maximize scalability and throughput-per-unit.
The outputs of the system will be power, fresh water, building material, and packets of recovered molecules (the ones that aren't made of H C or O) to hopefully be reused.
Glue + membranes + geometry + pressure = An open-ended variety of light-weight rigid structures with a very easy construction method. [5]
[1] https://en.wikipedia.org/wiki/Molten_salt_oxidation https://en.wikipedia.org/wiki/Molten_salt_oxidation
[2] https://en.wikipedia.org/wiki/Synthesis_gas https://en.wikipedia.org/wiki/Synthesis_gas
[3] https://en.wikipedia.org/wiki/Spittle_bug https://en.wikipedia.org/wiki/Spittle_bug
[4] https://en.wikipedia.org/wiki/Hydras https://en.wikipedia.org/wiki/Hydras
[5] https://en.wikipedia.org/wiki/Inflatable_air_cushion https://en.wikipedia.org/wiki/Inflatable_air_cushion & https://en.wikipedia.org/wiki/Air-supported_structure https://en.wikipedia.org/wiki/Air-supported_structure
- Jedi72 9y agoYou get an upvote for the sheer audacity of this project. Hope you make good progress, good luck.
- carapace 8y agoCheers! It's been slow going, working mostly in my spare time, but I'm committed. I've got two boats now for hosting the initial "kernel", and some time this summer to double down. One benefit of taking so long is that neural networks et. al. have taken off in the meantime and that should solve a few problems that might otherwise have been hairy.