3 ms·

This has the potential to do for inland transportation what containers did for long-distance haulage. The problem is the infrastructure. There's a lot to do b

This has the potential to do for inland transportation what containers did for long-distance haulage. The problem is the infrastructure. There's a lot to do before the system works at all. To bootstrap, some "community" needs to decide they want it, and then they need to build a small but extensible version. Extensions need to be able to "interface" with the existing system without necessarily being limited by it. I can see this working as, say, the main means of transporting physical materials around a new Olympics venue, or a new commercial district. It's basically an automated high-speed rail system, but underground.

jerf 16y ago
'Extensions need to be able to "interface" with the existing system without necessarily being limited by it.' This is the blocker. The way I see this happening is a robotic inner-city courier that can live with humans on a sidewalk. The feasibility of this is no longer pie-in-the-sky, now it's merely "on the horizon". Once you have this, it becomes actually economical in another couple of years. Once it becomes economical, there are too many of them on the sidewalk and you start considering things like giving them their own lanes, or their own paths through buildings, or, eventually, their own access to underground tunnels. From there somebody works out that they can make money building a tunnel from Manhattan to the rest of the city, and the network starts forming. The advantage of this approach is that once you have such robots you can start using current systems. Put a dedicated car on an otherwise normal freight train for these things, for instance. Eventually you've built a hybrid system that can accommodate all sorts of approaches. It ends up looking like the Internet and will probably recapitulate its development. There's no point stringing a tube across the country from one endpoint to another unless you've got a network already working on both ends. Attempts to bootstrap will generally fail because they will be too focused on top-down deployments of massive tubes that won't be useful for years; the winners will be the people who successfully bootstrap from the bottom and build the network from the bottom up. The tech isn't quite there yet, though, but getting close. So, I'd actually say it isn't even that "extensions" need to interface with the system, it is the system that will eventually be interfacing with the "extensions". The flexible robots have to come first, then we will build the tube network. Until those robots exist the tube network economics just make no sense, not even in a dense city.

jerf 16y ago

'Extensions need to be able to "interface" with the existing system without necessarily being limited by it.' This is the blocker. The way I see this happening is a robotic inner-city courier that can live with humans on a sidewalk. The feasibility of this is no longer pie-in-the-sky, now it's merely "on the horizon". Once you have this, it becomes actually economical in another couple of years. Once it becomes economical, there are too many of them on the sidewalk and you start considering things like giving them their own lanes, or their own paths through buildings, or, eventually, their own access to underground tunnels. From there somebody works out that they can make money building a tunnel from Manhattan to the rest of the city, and the network starts forming. The advantage of this approach is that once you have such robots you can start using current systems. Put a dedicated car on an otherwise normal freight train for these things, for instance. Eventually you've built a hybrid system that can accommodate all sorts of approaches. It ends up looking like the Internet and will probably recapitulate its development. There's no point stringing a tube across the country from one endpoint to another unless you've got a network already working on both ends. Attempts to bootstrap will generally fail because they will be too focused on top-down deployments of massive tubes that won't be useful for years; the winners will be the people who successfully bootstrap from the bottom and build the network from the bottom up. The tech isn't quite there yet, though, but getting close. So, I'd actually say it isn't even that "extensions" need to interface with the system, it is the system that will eventually be interfacing with the "extensions". The flexible robots have to come first, then we will build the tube network. Until those robots exist the tube network economics just make no sense, not even in a dense city.