So in this post about mass transit I made the same mistake a lot of amateurs do – I tried to solve the wrong problem. Kinda. Let’s try and fix that.
As I learned during my epic three-weekend Planetizen binge getting myself re-AICPed, efficiency in mass transit isn’t really the problem. More specifically, agencies spend a lot more on labor costs than they do on capital costs. Building smaller buses won’t get more service to more people; leveraging drivers better will.
At this point I think a lot of people just say “automatic cars” and figure the problem will be solved in 10 years (well, except for all those unemployed bus drivers). I used to think that, but with the recent deaths caused by testing of automated cars I’m less sanguine. Driving is mostly easy, but sometimes really hard. Buses in particular have to deal with the really hard cases more often than other vehicles. They’re picking up and dropping off people all the time; cars might be blocking where they need to go; they’re in heavily trafficked areas, with pedestrians, bikes, skateboarders, cars, trucks… the list goes on. The upshot is this – I think it’s going to be a lot longer than people think before we can actually replace drivers with computers. So what do we do in the meantime?
Next Future Transportation has an interesting modular framework, but like everyone else, they’re going all-in on automated driving technology. Maybe there’s room for something more like a modular bus concept. Pods of four to six, self contained and electric, like Next Future proposes – but rather than driving themselves, they would need to be linked to a driver pod. The driver could navigate more difficult areas that a computer couldn’t handle, and when pods detach might even be able to wirelessly steer the pod the last few feet to its final destination.
So why bother with pods over a standard bus? There are a few major advantages:
- By attaching and detaching pods, buses can move more efficiently. Rather than waiting for everyone to get on & off, pods can be separated from the whole and the rest of the bus can move on while people are disembarking.
- Pods can be parked in locations and wait for passengers. This would be like bus stops now, except the pods could be boarded and people could wait inside them for the rest of the bus to arrive.
- By utilizing a central dispatch, buses could operate more flexibly than just following rigid routes. Computers aren’t there yet for driving, but they’re very good at network optimization tasks. A central dispatch algorithm could be constantly calculating the most efficient way to route thousands of passengers in hundreds of pods with hundreds of drivers. This would also let agencies change their optimal tradeoff between waiting times, staffing, capital expenditure, and route service in software.
The goal here is not to replace drivers, but to utilize them more efficiently so as to squeeze more and better service out of the same resources (or fewer). Ideally the pods would be linked flexibly so that they form “trains”, rather than a single rigid unit, because that would increase the number of pods that could be strung together; they could be programmed to precisely match the drivers’ path while connected, and to park themselves while disconnected (automatic parking is already a feature of high-end cars, so not out of reach).
It’s always worth remembering that busing is primarily a means of transportation for the poor and working class, so the system shouldn’t rely on smart phones. Pods should have simple built-in screens allowing people to select destinations and call for pickup via a simple interface, and accept cash (maybe) or transit cards (definitely).