Robocars have been on the mind of city and transportation planners for some time. They have been told (correctly) that robocars will cause major changes to how people move around cities, and require a great deal of rethinking when it comes to planning and transit. The old 20th century ideas are wrong — but when do they change and what will the right choices of the future be?
The answer, of course, is that nobody is certain. And now that some robocar projects have been scaled back or slipped their dates, more are wondering just how they decide what to do. What if the technology doesn’t come as quickly as expected, or doesn’t come for a long time in the form expected. As this article describes, transit planners have a quandary. Some are enthusiasts, some are skeptics. Some are focused on their favorite solutions rather than on the goals that travelers, residents and governments have. They want to plan cities and infrastructure, but such efforts involve decades from planning to delivery. How do you plan when a plan from 2020 is almost certainly deeply wrong by 2030?
The answer of how to plan comes from the one industry that has had to live with constant change in the world under it’s feet — the computer industry. For over 50 years, computer chips have doubled in value every 1-2 years. No other industry has to deal with anything remotely like that. Imagine planning cities and transit if vehicles or construction or fuel cost 1/1,000th the price every 15 years, and kept doing that, decade after decade.
That industry’s answer has been to keep infrastructure as simple as possible, and expect to replace it frequently, putting as much value as possible into software rather than hardware. Software changes can be deployed for free, though they are of course not easy or free to create. The internet uses the same base design it had in the 1980s, known as the “stupid network.” All the smarts of the internet lie not in the network but in the endpoints — phones, laptops, web servers. They constantly change and improve and get replaced while the base network is as stupid as can be and stays the same. When it comes to infrastructure, you want to avoid long term changes because you know they will be wrong. Avoid them as long as you can.
There are analogs of this in the physical world of cities and transportation. Bare concrete is simple. It can carry cars, trucks, pedestrians, cyclists, motorcycles, buses, vans and vehicles yet to be invented. You can change the directions of roads, allocation of roads and rules of the road all with “software.” Even road signs and lights are “too smart” and should be replaced with software devices that let you change your mind. Here are some policies that can future-proof the urban design of the future:
BRT vs. LRT and other rail
Many cities love light capacity rail and related technologies. But rails, while slightly more efficient and able to let the vehicle steer itself, constrain you too much. Unlike pavement, rails can carry trains, trains, trains or trains. Modern tech can let you run tire-based vehicles as though they are on virtual rails. It’s quieter and a lot cheaper, too.
This has attracted many cities to BRT (Bus Rapid Transit.) Instead of a light rail or subway, they allocate a dedicated bus lane, sometimes with occasional dedicated bridges and tunnels. It runs more like a rapid transit line with stations — no paying on the bus to slow everybody down. Cities like BRT because it’s vastly cheaper, but its real virtue is that the infrastructure is simple. You can change your mind in the future, allocating the path to robocar services, regular shared use, or something new. You can even lay down rails if that turns out to be the answer.
The same philosophy applies underground. Boston’s “T” Silver Line is BRT which gets a tunnel but is able to emerge to the surface and use regular roads on the way to the airport, which would have been difficult on rails.
To future-proof even more, BRT “stations” must be offline. That means a bus can stop in one without blocking the path, allowing other vehicles to pass. Offline stations are hard with rail so old-school thinking doesn’t include them.
The bad news is that offline stations are hard, particularly for those trying to adapt a BRT lane into an existing road. That’s even hard with online stations as the platforms take away an extra lane during their length, and offline platforms would take at least two extra lanes, and ideally even more. If adapting existing tunnels with narrow stations it’s also challenging. If you have money, offline stations can be built above or below ground. Of course, above and below ground platforms require passengers to change level, which is not as convenient or accessible. Having the main path go above or below makes the pass-through ride less comfortable. For low volume lines, vehicles can pass by moving into the oncoming lane, which can be safe to do in computerized system.
Contrary to intuition, smaller vehicles can be more efficient than large ones, because they provide better, more frequent and direct service, which is much more attractive to riders — increasing the load factor and thus overall system efficiency. That’s why the A380 may never fly again. In the past no system could afford all the drivers this would take, but eventually that won’t be the issue, so be ready. Even if you want to design for big vehicles today, make a plan that can adapt to using smaller vehicles and different types of vehicles in the future.
This also applies to tunnels. While many wonder if Elon Musk’s “Boring Company” dream of cheap tunneling will come to pass, one thing is certainly true — smaller tunnels are already easier and cheaper to dig than big ones.
Smaller vehicles also make it easier to make offline stations.
Expect the Smartphone
You don’t have to tell people to expect the smartphone, one would hope, but people still don’t really think about a world where everybody, even the homeless have them, and they all naturally expect to use them as part of their travels and life in the city. We’re now at the age where if you don’t expect a smartphone in every car, you’re doing it wrong. Every day, more and more drivers never do a long drive, even their daily commute, without asking and obeying programs like Waze, like we’re little robots. In cities that have live transit data, you’re nuts to use a bus or sometimes even the subway without consulting it, and people know it. We’re entering a world where the way people use the infrastructure is under the control of software, and on the cusp of being able to insist on it. Our early experiments with congestion charging, managed lanes, ramp meters and coordinated traffic signals are just the beginning.
Consider a world where cities can give advice and even orders to cars to avoid congestion. Where traffic volumes can be set to make streets peaceful where we want them and serving flow demand where we want that. The old idea that it’s a big network of streets that people just squeeze themselves into is already fading. Make a plan that’s ready for that, while also ready for the pace of change to be unknown. Cities know not to install dumb traffic signals, and you do want all of them to be software devices, but the reality is that most drivers should rarely encounter a red light if you do it right, because a higher level software infrastructure doesn’t wastefully advise them to drive in a way that encounters them.
The rule of the smartphone is simple: The smartphone always wins. If it can be done in a phone, it’s a losing battle to do it any other way. Many efforts at building a “smart city” have been, and will be obsoleted by technology in smartphones. The reason is straightforward — people replace smartphones every 1-2 years, and because they are sold by the billion, they have vastly superior economies of scale over any product. Your smartphone is only a year old, and was designed just a year before that. The software on it was updated yesterday. That’s unlike everything else from infrastructure or industry. Even if the smartphone doesn’t seem like the leading candidate for an application, it will be.
A great example can be seen in the measurement of traffic in cities. For many years, cities were eager to measure what was happening on their streets. They spend millions digging up streets and installing sensors and cameras and systems. At the same time, a few folks in Tel Aviv built a smartphone app called Waze. Without spending a dime on infrastructure they quickly had superior traffic and movement data on almost every city in the world.
The internet principle of putting the intelligence at the endpoints is at play here. Don’t make the smart city — rely on smart vehicles and smartphones in the hands of the people.