I’ve been listening to some really smart people talk about the future of cities in the past few weeks. Although they all treat different themes from different perspectives, they are pretty consistent about how badly we anticipated today’s needs back when there was still time to lay out our cities differently. Underlying their scolding is the idea that we should have employed design thinking in our long term decision making and that the opportunity exists now to correct past failings.
A good example is Dr. Kevin Stolarick’s argument that Toronto’s 21st century transit woes are the result of errors in planning and zoning. In short, transit problems are actually housing problems.
He contends that we built housing too far from places of employment, making it necessary to handle huge volumes of people going from one place to the other. Had we mixed commercial and residential development more effectively, enabling more people to walk or cycle to work, we might not be spending billions of dollars to bring people from suburbs to the core and back every day, in numbers that overwhelm subway and surface systems at peak times.
His point is not only to chastise us for past sins, but it’s also to caution us against repeating the error. Until we recognize transit problems as housing problems, we’ll screw up both housing and transit again, and, decades from now, another generation will be scratching their heads about why they pay so much tax and waste so much time on transit.
I can’t disagree with either point. Public transit compensates for the layout of cities in which few people can live where they work. It’s a necessary evil that we should make less necessary as we plan for the future.
No one wants a big, elaborate, costly transit system, do they? People don’t line up for thrilling subway and bus rides on summer weekends, like Ferris wheels or roller coasters. They eat up our time and money. They are an alternative to something worse, but are no joy at the best of times.
Nor are they hugely virtuous. Although you can’t see the emissions from electrified subways and surface rail, all that power is being generated somewhere out of sight, along with the toxic waste of energy production. Then there are the financial and environmental costs of building and maintaining transit infrastructure, much of which is idle most of the time. Public transit may be preferable to car travel, where supported by density, but in the final analysis, we could do with less of both.
Which brings me to a quandary raised by Terry O’Reilly at The Walrus Live (on which I recently blogged) and by Gabe Klein at the Building a Better City conference in Ottawa. Both acknowledge the inevitability of autonomous vehicles, yet neither advocate for a coherent planning response to the dilemma of investments in roads and rails.
I suppose that’s not their responsibility. Yet, an incomplete vision of the future isn’t very compelling. Any city that’s planning to add surface rail must consider the fact that demand may diminish abruptly as autonomous vehicles take over our roadways. All predictions are that this process will be complete in a matter of decades, well before the full value of any new urban rail system can be realized.
We are accustomed to viewing private and public transit infrastructure as a zero sum game when we budget and plan for the future. In Toronto we have lately come to the realization that investments in roads have been forestalling necessary investments in rail and other forms of public transit. Other cities are making huge investments in light and heavy rail now. But does this still make sense?
Autonomous vehicles blur the line between public and private transportation in ways that should make us rethink the priority we’re currently putting on “public” systems. Although roads currently serve the needs of private commuters and rails carry public transit vehicles, this won’t be a meaningful distinction once car ownership ceases to be the norm and rubber-tired, autonomous vehicles become capable of linking up with software and sensors to function more like trains.
Car ownership becomes unimportant when there are vehicles waiting nearby to be summoned by phone and are stored, recharged, and maintained for you somewhere else, by someone else, for the 97% of the time that an average car is currently not in use. Vehicles will be owned by fleet operators, possibly by the manufacturers and will be insured as the liability of their owners, operators, or software developers, not the riders. The savings in space, time, and money will eventually make this result irresistible.
When I refer to the close linkage of vehicles by software and sensors, I’m referring to the way vehicles can drive once human error has been eliminated. When speeds are uniform, the gaps between cars can be reduced. Lines of traffic can be shortened and lanes can be narrowed. Cars can be parked in much tighter spaces once you get humans out the equation.
What essential difference is there between rail transit and a line of autonomous vehicles driving on pavement? Both are chains of wheeled vehicles with many passengers and one or fewer drivers, and both are better than private vehicles on today’s roads, in which drivers outnumber passengers. The important question is, however, whether or not “public” rail transit is inherently preferable from a planning perspective.
The advantage of autonomous vehicles, be they buses, sedans, or two seaters, is that they’re linked by software and sensors, rather than by the physical couplings linking LRT cars, articulated street cars, or subway trains. The links in a chain of autonomous vehicles will be free to disconnect and follow their own route at any point in their trip, whereas the rail transit vehicles are trapped on routes dictated by where the rails and power lines were laid. If you want to deviate from the transit system’s route, you have to leave the vehicle and find another way to get where you’re going.
That’s an unnecessary limitation given that autonomous vehicles can run on simple pavement and can approximate the speed and volume of subterranean or surface rail because they take the traveller from their point of origin to a precise destination. Rail vehicles have to run where the rails go, which at best brings the majority of riders somewhere near to where they’re trying to go, or at least to a transit connection that will bring them closer. Add the walking time on both ends of the trip to the connections between routes, plus the time actually in motion, and the average speeds of door to door autonomous vehicle trips could compare favourably with the experience of most public transit users.
When comparing modes of transit, there is also the question of efficiency. Public transit is designed to meet peak demand times, which occur at the start and end of work days. With the exception of major public events, streetcars, buses, and subways are mostly empty most of the time. There’s nothing that can be done about that except to reduce frequency of service, which damages perception of the system and hurts ridership.
Autonomous vehicles can be designed to avoid some of this waste. Just as with Uber today, riders can summon a private or shared vehicle, and the vehicle can be selected by the system according to the number of people and their routes. Compared with the minimum standard sizes of todays streetcars, subway cars, and buses, utilization rates will rise as the frequency and size of vehicles is matched more closely to the specific requirements of their routes and riders.
I’m actually not trying to make the case that one is better than the other. I’m trying to point out that you can’t optimize both on the same thoroughfares. Despite general agreement that autonomous vehicles are the future, we continue to make investments in rail transit. If it’s reasonable to imagine that it will take 30 years to achieve full automation of vehicles on major urban thoroughfares, parallel investment in rail infrastructure, designed to last 50 years, will be redundant. Imagine a lot of mothballed rail vehicles in a transit yard somewhere while streams of autonomous vehicles move quickly and efficiently on rubber tires over kilometre after kilometre of useless steel rail.
What’s wrong with this picture? It’s that the dilemma of investments in roads vs. rails, left unresolved, will retard the development of either system to its full potential. Decades will pass while politicians dither over sunk costs in rail transit, while frustrated autonomous vehicle advocates and planners battle unnecessary delays in the infrastructure needed to achieve the safety and efficiency goals of smart roads and vehicles.
So, while it’s good that we have two enlightened options to the current mess of private cars and errant drivers jamming our roads and polluting our air, we won’t fully enjoy the benefits of either one unless we choose between them. That priority for investment needs to be made now so that competing demands on the public purse and road space doesn’t result in catastrophic waste and delay.
Listen hard to the chatter about the future from academic oracles, planners and politicians, and you won’t hear doubts or detractions about the value of rail transit. You’ll hear lots about autonomous vehicles in the future, but little about the hard choices that presents us right now. The technology needed for autonomous vehicle road systems is already available, but our thinking is still stuck in the 20th century debate about the prioritization of public transit over private vehicles.
That debate will soon be irrelevant, if it isn’t already. “The war on cars,” at least in Toronto, is an anachronism. It has already been won. Now it’s time to for planners, politicians, and the public to contemplate what form of peace we’d like. It’s not for futurologists to contemplate in lectures and colloquia. Clear and coherent decisions are needed now. Time has overtaken the transit debate. The future starts now.