Right. When charging range gets consistently underestimated in the real world vs. reference/test specs and the shortfalls hit seemingly similar route networks harder than others (e.g. Shenzhen being a mild-weather city but performing more poorly on fleet uptime than same makes/models deployed elsewhere)...then there's still big unknowns out there that haven't been identified. You can't just cite "of course a vehicle rated range of X miles, plugged into this route of ours at X - Y miles, with one test bus that made mincemeat of extremely small samples sizes on one route" as justification when that's exactly the assumption those other cities made before impaling themselves.
I read in one of the link-outs on an article about Shenzhen's and Moscow's problems that the electricity budget on the average modern traction-power (TT, battery, or hybrid) bus goes only 20% to propulsion, 40% to HVAC, and 40% to "all other" such as lighting, 'kneeling' hydraulics, cameras/sensors, ASA, radios, etc. Sticking a miles-of-range reference quote on it literally is much less than half the story on where that energy budget gets apportioned during the course of the trip, and that's how the planners of Shenzhen--who certainly didn't set out on this adventure to look like dummies--ended up whiffing so badly on their range targets.
The HVAC figure didn't shock me at all; in fact, heating takes more power in an electric bus than a diesel because cabin ambient temperature isn't getting any boost from waste heat trapped off the engine by radiators. That, in addition to the physical degradation of the battery's ability to take/retain a charge in cold and/or cold + wet weather, is why the HVAC budget looms so incredibly large. But I was floored that all the other electronic doodads gobble up such a massive share of the budget. There's basically a whole server room stuffed in there for the computer controls and sensor plant, stuff that's always-on and living up to computers' and routers' reps as electricity guzzlers. Don't anyone get too precious about demanding onboard WiFi in the next battery bus fleet in the name of Jetsons Shit-begetting-Jetsons Shit frills; that's going to murder the energy budget a little bit more. Quantity of stops probably has a big effect as well and opens up stark disparities in routes where the kneeling mechanism has to be used often vs. rarely. Dedicated infrastructure like the express-oriented Silver Line is probably going to fare better than a dense city bus route once electricity chew on kneelings is tallied up, but the number of routes a test bus would have to be cycled on throughout the district to establish those ranges is huge before they have enough actionable data to make good decisions. One Transitway test bus tells them almost nothing about what it takes to convert the whole damn district over to batteries. And some individual routes are going to fare better than others when Boston traffic congestion characteristics are factored, because reference charging ranges are based on some league-average of regenerative braking charges taken en route. Well, what kind of braking maneuvers do you have to do over the course of a trip through city streets to hit the 'reference' range? Do you get enough charge-lets sitting in stop-and-go traffic downtown the whole time, or does that charging energy need to pile up braking by blocks/traffic lights at 15 MPH instead of braking in 10 ft. increments by the taillights of the cars in front at 5 MPH? Some routes in Greater Boston are going to present wildly different energy recapture because of that. What about hills?...do the routes that have some easy recapture coasting downhill on a brake application fare better in traffic than those running at flat elevation?
Don't get me wrong, these questions are getting backfilled with answers at a pretty good clip. But I don't see how anyone can take a look at the energy budget for that type of vehicle, tally up all the ways the nooks and crannies of one city's route network--route-by-route--gobbles that energy budget at divergently different rates, and conclude "MOAR IMPATIENCE!" as the answer. There's no practical choice but to slow-walk the adoption, because that's the only way to sniff out all the network-specific asterisks that impaled Shenzhen and Moscow. Work smarter, not faster, when it comes to planning this adoption out.