But buses have a large production volume to spread out the R&D spending, and the fleets are regularly refreshed (since buses are basically driven into the ground over 12 years.) For once-in-a-generation replacement of the ~260 streetcars and other LRVs that Toronto has, it'll take much longer before our existing fleet is replaced with autonomous trams (and there will probably be a larger price premium for it.)
I'll go out on a limb here and suggest that retrofitting our LFLRVs with autonomous tech might not be too difficult once the autonomous technology is available on the market. As long as the LRVs are
fly-by-wire, it should be pretty straightforward to connect an autonomous driving system to the controls of the LRV.
The more difficult parts might actually the upgrades that would be needed elsewhere in the network. For example, all the network switches would need to be electronic, as they'll be no drivers available to manually move the switch. This won't be an issue on the Crosstown, but it will be an issue on the legacy streetcar network.
Electronic communications between Transit Control and the LRVs would need to be improved as well, as failed communications system would be catastrophic under autonomous operations. I'd image that some combination of cellular networks (4G, 5G, etc...) and
mesh networking (essentially the streetcars wirelessly relaying messages between each other) would be sufficiently reliable. These solutions are proposed for use in autonomous automobiles as well.
It's exciting to think about the service improvements that would be delivered by autonomy. Operating costs would be significantly lower, which means we can afford to run additional trams. And without having to manage people, you'd no longer have to deal with layovers to switch drivers.