AeroTrains

It's a branding thing. Aerotrains as a generic term for 'AirTrain' or 'SkyTrains', where you've got fixed guideway rail to the airport. Fixed guideway rail to the airport is a bit different from having a people mover on the airport. People Movers (whether rubber tired or steel wheeled) are on-airport trams, on separated guideway, typically elevated and automated. They typically run between the terminals and the rental car lots. Dallas Fort Worth (DFW) and Detroit (DTW)  are the ones I'm familiar with. Wikipedia lists a whole series of them, but it's a mixed list that includes things that aren't really people movers. It is an interesting mix of systems inside the secure corridor, and outside the secure corridor. Clearly, a train can't cross that barrier, so we can disregard a direct connection with such systems and the broader rail network.  So we're really thinking about a system that servers the terminals, the rental car centers, and short/long-term parking. Ie,  'train as shuttle'. Which is a silly way to use most trains, as it underuses their capacity. You don't want vehicles capable of 70 mph noodling about at 15 mph--it's like always driving a Mustang at 30. The trains don't have the proper operating specs for shuttle use. For a shuttle, you want a light-weight vehicle, standing room only, with rapid acceleration. And automation. So we've established that using a commuter train (light rail, heavy rail) for intra-airport circulation is silly.

It's also hard to get a proper train onto the airport. Ticklish bit of Small Starts is that the Feds (reasonably) expect a transit agency to own the right of way under their tracks. (Longterm leases are sometimes accepted). After all, no one wants some quisling mayor refusing to renew the lease, and disrupting a critical link in the regional rail network. However (understandably), airports don't like having someone else own part of their airport. And so you get nonsense like LAX, where the light rail stops off the end of the airport.

So when we talk about an Aerotrain, what we're really talking about is a train to the airport terminals. To the curbside location where drop-off and pick-ups occur.

Not to say that people movers aren't cool--but they provide the kind of 'magic carpet' transit shuttles and buses are (irrationally) expected to provide. It's feasible, but spendy. It requires grade-separated guideway, and that's never cheap. Grade-separated guideway is required for automation, and automation in turn is required to break the link between labor costs and transit service costs.

It can be argued that advances in automated vehicle technology make mixed traffic (or partially mixed traffic) operations feasible. (Or at least time-separation). I'm uncertain. Some of the people movers run a vehicle every 30 seconds. So every 15 seconds, the crossing would close as the 'gates' come down, and people can't cross that line. Hardly impossible, but how much carnage would an airport be willing to tolerate? One fatal crash per year? IIRC, value of a human life is running about $2.x million. Likely higher for airline passengers, who tend to be more affluent.

If you make it safe enough to preclude that, then you have the potential problem of lost child/argumentative grandmother standing on the guideway, as automated vehicles politely wait for them to move, with a queue growing behind them. So it would probably require an employee to stand there (in white gloves) and shoo people out of the guideway.

So it looks like any airport requires three different fixed guideway systems: A) inside the security cordon, for inter-terminal circulation, B) for land-side circulation, between the terminals and parking lots, and C) to connect the airport to the broad metropolitan area.

Option A is only worthwhile if the size of the airport is such that it exceeds that which can be serviced by moving walkways. They seem to be quite common. Phoenix (PHX) has reached the point where they use two-speed walkways, with lower speeds ones jumping onto higher speed ones. (One always seems to be out of service, someplace in the airport, making me question their reliability).

We can't combine B with C: Different operating characteristics, different geometric requirements, etc. So an airport requires a B system to connect to the C system. It needs very high headway, so it needs to be automated, so it needs to be grade separated, so it's going to be expensive. Anything that's not automated, you might as well just stick to using shuttle buses. Although that does suggest an interesting niche for an Automated Transit Vehicle System (AVTO). While they aren't feasible for real-world operations, they might be feasible for pedestrian free vehicle-only 'campus' environments, like airport parking lots/garages/rental car counters. It would be an automated solution that doesn't require grade separation, so it would be much much cheaper.

So a B system provides on-airport circulation. Could it be carried off-airport? Probably, but not very far. Remember that it would likely be a shuttle system with expensive grade-separated guideway, so it would be costly to build very much of it. Such a system could follow the 'pre-metro' paradigm, where only parts of the line are grade-separated. Such a system should not go too far beyond the operational envelope of the 'shuttle' portion of the line, so it would have a limited top-speed.

But that would also generate an 'agency issue'. Why would an airport (the necessary owner of such a system) do such a thing? Extending the network beyond airport property is clearly fraught with inter-agency conflicts. Why extend their system beyond their service area? Empirically, it only happens when they need to cross a 'moat' of property they don't own (a freeway, a roadway) to access airport property on the far side of the road ie: Atlanta (ATL). Surprisingly, they also have a hotel, something I had thought unique to Barcelona. It does make me wonder if such a 'B' system could be the basis for an Airport City.

But, essentially, the role of B systems seems to be intermodal: Connecting airplane passengers to surface transport vehicular modes: cars and trains. Atlanta's system meets that, connecting to MARTA. And that puts the JFK AirTrain in a different context. The JFK AirTrain (map) is interesting in that it doesn't just reach a terminal station, but reaches a transit hub, with connections to multiple lines and multiple systems: The NYC Metro and the Long Island Rail Road. Which is nice, because it cuts out the need to make an inter-line transfer to reach the airport.  Off the cuff I'd estimate it's maybe 12,000 feet from the airport to Jamaica (Wiki says 3.1 miles guideway). Funding seems to have been the breakpoint: The system funding source was an airport passenger fee, which could only be used to benefit air traffic users.

Reading up on the SkyLink at DFW makes me wonder if the A & B systems could be the same. Clearly not the same guideway (one being within the cordon, one being outside it), but possibly the same type of guideway and vehicles. It would certainly simplify maintenance requirements, and allow A system vehicles to be garaged/maintained/repaired off airport property. Only a limited (gated) interlink would be required.