Self-Driving Cars Are Coming Sooner Than We Think
Cities everywhere should be figuring out now what it means for transportation and land use.
Promises, promises
Since the 1950s, we have been promised that self-driving cars are about a decade away.
By 2013, Elon Musk was promising self-driving cars within two years. In 2014, he was saying by the year-end. Since then, every year he continues to reiterate some variation on this promise.
Personally, I had given up on seeing self-driving cars anytime soon. But a recent experience has caused me to do a u-turn.
Let me share why I think self-driving cars are coming sooner than we might expect. And what that could mean for cities.
California dreamin’
I was in California last week, and had the chance to ride in a Waymo self-driving taxi.
Waymo taxis are fully autonomous vehicles. The driver’s seat is empty. The steering wheel turns itself.
You call a car using an app. The experience is much like an Uber, but without the need to leave a tip.
Waymos have only been available to the general public in San Francisco for less than 50 days now. They are also available in Phoenix and will be coming to Los Angeles and Austin later this year.
The cars are trained to be risk-averse. They are cautious, make good driving decisions and — unlike many human drivers — aren’t continually checking a phone.
Riding in a self-driving car felt remarkably safe. In fact, driving in an autonomous Waymo vehicle felt much safer than many Uber or taxi rides I’ve taken.
What about a Canadian snow storm?
The technology behind self-driving cars is still in its infancy. Cruise, a rival operation from General Motors, pulled its fleet of self-driving cars off the road in late 2023 after one of its cars crashed into a pedestrian.
And of course there are limitations. For now, Waymo taxis operate in pre-defined areas that the cars know well.
And they operate in cities with mild winters.
Will fully autonomous driving ever be possible in the middle of a major snow storm? Who knows.
But if self-driving cars can work in most conditions, expect to seem them on the roads when conditions allow. A few bad snow days will not become a deal breaker for the rest of the year.
Growth looks like a hockey stick
The presence of self-driving public taxis on public roads represents a major milestone. Whether we like it or not, self-driving cars have officially arrived.
The growth of autonomous vehicles is taking on the shape of a hockey stick. We’ve been in the experimenting and testing stage, where growth is flat and looks like the blade of the stick.
But with self-driving cars now accessible to the general public, we’ve reached an inflection point. We’re moving from the blade of the stick to the shaft.
I expect self-driving cars to make an appearance in Canadian cities within three-to-five years. Taxis will be the first use case, in advance of more widespread adoption.
I now believe we are now talking years, not decades, for self-driving cars to become commonplace.
Considerations for cities
Cities will want to start thinking through the implications of self-driving cars. Here are five starting points to consider.
1. Congestion
One near-term outcome of self-driving cars is that we simply end up with more cars driving around aimlessly on our streets clogging up traffic.
Maybe even turning today’s congestion into tomorrow’s gridlock.
With self-driving cars, we will be faced with a choice. Suffer through increasingly bad traffic. Or address congestion head-on, through regulation or road pricing.
Cities should now be trying to forecast the impact of self-driving cars on traffic volumes, and raising public awareness about the options for managing congestion.
2. Parking
So many land use decisions in cities are driven by parking concerns. Self-driving cars will be an important impetus to examine past decisions around parking.
Self-driving cars don’t require prime parking spots. Self-driving cars need to be parked at some point, but that can be at a more distant location. Parking lots in choice locations can be repurposed for higher value uses.
Many cities still have minimum parking requirements associated with different buildings and land uses. Minimum parking requirements make little sense today. They will make even less sense tomorrow in a world of autonomous vehicles.
Cities should be relaxing those parking minimums now. Any plans for new parking lots in prime areas should be put on ice.
3. Housing
The federal government is pressuring Canadian cities to change their residential zoning bylaws to allow 4 housing units on any residential lot. But upzoning alone is unlikely to create the supply response desired, because it’s very hard to fit 4 housing units on a lot if you also need to provide parking for some or all of those units.
If parking is moved offsite, it is quite manageable to fit more than four units (zoning permitting) on most single lots. With self-driving cars, parking could be relocated to centralized parking lots a few minutes drive away (again, zoning permitting).
4. Transit
Some people may see self-driving cars as an existential threat to transit. I don’t agree.
Transit is about the mass movement of people. Transit gets a large number of people into a concentrated area, by buses or trains. Road capacity can’t manage everyone in a car — regardless of self-driving or not.
Rather, I see two implications of self-driving cars for transit.
First, buses themselves will eventually become self-driving, which would suggest a smaller workforce and presumably lower transit operational costs.
Second, more transit agencies are replacing fixed routes with on-demand transportation, for low-density areas, first and last kilometre service, and underperforming routes. Autonomous vehicles may become a viable option for offering these on-demand services.
5. Vulnerable road users
Cities have a special responsibility to vulnerable road users as they begin the transition to self-driving cars. Autonomous cars can identify other cars or trucks without much difficulty. Identifying pedestrians and cyclists is considerably harder.
Transitions are never easy. Pedestrians and cyclists should not have to be the bloody face of the transition to self-driving cars. Cities can mitigate the risk of autonomous cars to vulnerable road users by building out the safe cycling networks and protected pedestrian crossings that keep different modes of travel seperate from one another.
Not a silver bullet
Self-driving cars are not some sort of panacea for cities. I would much rather see cities provide real options that allow people to get around town without having to get into a car.
But I have experienced first-hand a highly-compelling new technology. A technology that has the ability to reshape cities for better or for worse.
Self-driving cars are now on the streets of San Francisco. I see them spreading quickly to other cities; as quickly as regulatory frameworks allow.
Cities can wait for self-driving cars to arrive and then react to the growing congestion on their streets.
Or cities can start preparing for a world with autonomous vehicles. I’ve listed five areas for cities to consider as they look ahead to a world with self-driving cars.
If you have other thoughts on how self-driving cars might impact cities, please share in a comment.
Thanks for starting this important conversation. We don’t think about this enough!
Implicit in your analysis I think is that, alongside the shift to driverless vehicles we we also see a shift from personal ownership to more common use of shared vehicles as a service to get around. Given cost considerations, I think that’s likely. Driverless vehicles will also blur the lines between car-sharing and ride-sharing or taxi services.
We were in San Francisco a few weeks ago and used Waymo to go to and from a restaurant. We also used it two years ago in Phoenix, but I think the technology has improved since. I agree it's promising but I don't have a sense of costs relative to conventional taxis.
Three points:
1. I don't understand why self-driving cars would lead to more congestion. That implies that people would take additional trips or shift from other modes such as transit, which woukd require a cost much lower than current taxis or a service that is significantly more convenient
2. As an older person, I look forward to a service that would enable me to retain more mobility and autonomy
3. As a cyclist and pedestrian, I agree we need to be concerned about vulnerable road users, but I suspect that self-driving cars will reach the point where they are much better at detecting and avoiding cyclists and pedestrians than many human drivers