geospatial

I'll share more as the thinking develops, but I would love anyone to ask questions, point out flaws, sharpen my theoretical thinking etc 🙏

Here's a proposal I put together for the precision congestion systems using specialized, context-aware bonding curves:

Here's a proposal I put together for the precision congestion systems using specialized, context-aware bonding curves: 

https://github.com/johnx25bd/network-load-balancing

The crux of it all (besides a million things including proof of location, usability / adoption etc) is the pricing model. 

Prices to access network segments will need to be affected by local area network conditions (in both time and space), as well as agent (i.e. vehicle) characteristics.

... as well as network administrators priorities — do they want to optimize for reduced congestion? Transport times? Network emissions? Road wear and tear? Routing around some emergency? etc.

I'm pretty early on in thinking through how to create this pricing model, but it seems like it could be fruitful to train an artificial neural network on loads of simulations (and possibly actual network usage data) so it can produce prices that incentivize the network actors to choose routes that results in the ideal network configuration.

Anyway, I'm having fun thinking this through, and am very (very!) open to feedback, critique etc.

I decided to de-emphasize the onchain element for this proposal (sending it to some friends who aren't see deep into Web3), but I see the market infrastructure for this being on-chain.

This would require the deployment of hundreds or thousands of tiny markets for time-bound network access tokens. No better way to do that than with automated market makers.