I imagine that in terms of capital costs buying AGT-capable trains probably at least pays for itself, as station costs tend to dominate subway construction. Because AGT trains can be run at half length and double frequency at no additional operating cost, the cost savings from using smaller station boxes often seem to pay for or even outweigh the added expense of AGT signaling and software. Going the AGT route feels like a no-brainer.
I imagine that in terms of capital costs buying AGT-capable trains probably at least pays for itself, as station costs tend to dominate subway construction. Because AGT trains can be run at half length and double frequency at no additional operating cost, the cost savings from using smaller station boxes often seem to pay for or even outweigh the added expense of AGT signaling and software. Going the AGT route feels like a no-brainer.