There is too much pre-existing infrastructure underground to automate it completely in a cost-efficient way. In general, almost all construction work other than heavy lifting is still done by people.
"There is too much pre-existing infrastructure underground to automate it completely in a cost-efficient way today."
I've said on HN a couple of times that robotics work is going to radically affect society, and this is the sort of thing I'm thinking of. If we had real robots that could be turned loose with nothing more than a layout of the desired network and the necessary resources, and could intelligently deal with most blockages and intelligently call in human help when needed (and only when needed), then the cost equation of this idea changes a lot. (Also, robots to do a lot of basic maintenance without human intervention are probably necessary.)
Right now, it's a total pipe dream. Perhaps when we get such robots, it'll be more cost-effective to do something else instead, like an above-ground robotic bucket brigade for item deliveries. Or robotic delivery vehicles. But the tunnel approach may turn out to be feasible. (Or, more likely, a bucket brigade in some places, tunnels in others, and robotic drivers in yet others, where each makes most sense, freely intermixing delivery styles as needed.) So, bear in mind it is not that I am predicting specific results like "tubes in the ground", but that robots are going to radically change the cost side of many cost/benefit analyses.
"robots" (forever undefined beyond just that they're robotic) have been promised as the general fix-all to the world's physical labors since at least the 1950's.
Regardless of who does the labor, the reason robots have not spread further is that its still significantly cheaper to have people do things. So why spend the money replacing people with robots in this situation, all to build something that can be done already?
"general fix-all to the world's physical labors since at least the 1950's."
First, I'm not saying they're a "fix-all". I'm saying they're going to disrupt cost/benefit relations increasingly over the next few years. That's a much more defensible statement.
Second, part of the mental block that people are having with understanding this is that the promises that were made in the 1950s were premature, not wrong. Look at actual research coming out of robotics right now, not from the 1950s. It's still early, but there's been a qualitative change in the past couple of years; things are moving again. We got used to "nothing happening" so much so that it became just part of the mental background of our lives, but that doesn't make that rational.
There's nothing stopping people from being part of the delivery network, either. The most likely outcome is that it simply becomes an outgrowth of today's Fedex and UPS, starting in the cities and spreading out as the economics make sense. It's not even as much of a discontinuity as you might think, as from what I've seen of photos of the inside of the hubs, the hubs are already building-sized robots with small amounts of human help. It's just an expansion of the already-robotic-core of the delivery system, not an introduction of a brand new concept.
Agreed, the amount of work that would have to go into placing thousands of these tiny subways under the city to deliver to every building is insane, especially considering that they could flood or even a minor ground swell could dislodge the tracks causing a jam.
Sticking something like this underground just sounds stupid to me, especially when it would be impossible to have people actually go in to fix things. IMO this would either need to be built onto existing subway infrastructure where access to the system for repairs will be relatively easy . . . or do the cheap-easy way of just build it above ground.
I mean most large cities are getting expanding amounts of skyway (the enclosed walkways between buildings), once this gets to sufficient scale it would be no extra effort to attach pre-fabed tunnels to the bottom of the skyway networks. It's currently part of the Calgary by-law that all new buildings have to be connected to their skyway system.
The mass installation of new infrastructure is absurd considering the amount of work it requires below-ground, however above-ground would be exceptionally simple with a well developed skyway system. There's very little point of implementing this system without making it big, so until its exceptionally cheap per mile to install it will never happen, because our cities are simply way too big for a project like this if it's costly per mile.
